From 8da91e3754d5d5a57718be568abd8984ffbd584e Mon Sep 17 00:00:00 2001 From: Claude Date: Fri, 18 Sep 2026 20:51:46 +0000 Subject: [PATCH 01/21] =?UTF-8?q?spec(epic):=20W4=20=E2=80=94=20work=20rou?= =?UTF-8?q?ting=20&=20package=20cohesion=20(FIX-1407)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The epic-spec for W4 as four documents and three figures: the objective as teams before and after, what's in the box with the exit gate drawn as the top row, the live set table for all seven parented issues, the cross-cutting decision cards carried from the Architect's EM guidance, the rules every child obeys, and the path with the W3 ship fence. The set is written as seven, not the four the epic body names. FIX-817, FIX-1415 and FIX-1430 were parented later; they are recorded with the split visible and put to the objective gate as Open 1 rather than resolved on the branch. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01PxDE76gD3BxXnPPjT7BT1T --- .../BUSINESS-RULES.md | 62 +++++++ .../DECISIONS.md | 160 ++++++++++++++++++ .../work-routing-package-cohesion/PLAN.md | 101 +++++++++++ .../work-routing-package-cohesion/SPEC.md | 128 ++++++++++++++ .../figures/end-state.svg | 78 +++++++++ .../figures/ownership.svg | 107 ++++++++++++ .../figures/path.svg | 103 +++++++++++ 7 files changed, 739 insertions(+) create mode 100644 spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md create mode 100644 spec/_epics/work-routing-package-cohesion/DECISIONS.md create mode 100644 spec/_epics/work-routing-package-cohesion/PLAN.md create mode 100644 spec/_epics/work-routing-package-cohesion/SPEC.md create mode 100644 spec/_epics/work-routing-package-cohesion/figures/end-state.svg create mode 100644 spec/_epics/work-routing-package-cohesion/figures/ownership.svg create mode 100644 spec/_epics/work-routing-package-cohesion/figures/path.svg diff --git a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md new file mode 100644 index 0000000000..db0c8054a9 --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md @@ -0,0 +1,62 @@ +# FIX-1407 · Rules every issue in the set obeys + +[Spec](SPEC.md) · [Decisions](DECISIONS.md) · **Rules** · [Plan](PLAN.md) + +At epic altitude the rules aren't behaviours of one feature; they're the constraints every child +spec and implementation must satisfy, and the place a cross-spec review checks. Each says who owns +it and where it's checked. ER-1 to ER-5 and ER-19 are the rows of the ownership matrix in +[DECISIONS.md](DECISIONS.md#who-owns-what). + +## What a team gets, and what it doesn't + +| # | Rule | Owner | Checked at | +|---|---|---|---| +| ER-1 | Work for a team lives as a row on a **channel or org board** — `0..N` TaskCollections attached to a channel — and a seat claims it or is assigned it. Channel flow actions and `taskTools` are **two doors on one mutation surface** | FIX-1385 | FIX-1385's spec · the proof | +| ER-2 | A seat carries **one package format**, with two attachment modes: seat **always-on**, library **opt-in**. A tool, a skill and an instruction have one home between them, not two | FIX-1394, after POC ratify | FIX-1394's POC matrix · every child's spec review | +| ER-3 | Membership expand, author check, fan-out and DM find-or-create are answered by **one L2 query** over the shared Workforce map plus hire, `channelInstances` and open sessions | FIX-1405 | FIX-1405's two prove-pressure callers · FIX-817's and FIX-1415's spec review | +| ER-4 | Dispatch always passes `parentSessionId` and the child binds its parent **for life at mint**. Parent history reaches a child only through **opt-in tools**; the default payload is the brief — `goal` / `constraints` / `acceptance` / `links` | FIX-1408 (D4) — **shipped as a decision** | Every child's spec review | +| ER-5 | A board `assignee` is a **board-worker key**, not a Workforce seat; the two map by composition. `assignee` stays optional, and a registry board's "must be assigned" is discipline, not schema | FIX-1385 | FIX-1385's spec · FIX-1430's wiring | + +**On ER-2**: *after POC ratify* is the rule, not a hedge. FIX-1394 does not cut a ship ticket from +exploration alone, so no other child may write against a package schema before that ratify lands — +the schema itself is on the [Still open](DECISIONS.md#open) list. + +**On ER-4**: it is discharged by a decision rather than by a check, because FIX-1408 closed +without an implementation PR. The first child to build over the wire is where it first gets tested, +and its unclosed walls are [Open 2](DECISIONS.md#open). + +## What no child may do + +| # | Rule | Because | +|---|---|---| +| ER-6 | No child invents **Agent, Channel, Team, MessageBoard, TeamFlow, ChannelAdmin or SessionBoard as an L1 package type**, and none revives `materializeAgent` or an `AgentRegistry` | D3. Agent is an opinionated default *kind*; the rest are L2 opinion or nothing | +| ER-7 | No child ships `team.*` (or other) wildcards before the inventory lands, and none blocks the channels convention on them | D5. Wildcards are a later **caller** of inventory, not part of the first ship | +| ER-8 | No child ships the package collapse from FIX-1394 alone, without a POC ratify | The whole delivery posture of that ticket: POC matrix → Architect/PM ratify → then cut ship tickets | +| ER-9 | No child dumps a parent's full transcript, or a parent's private seat history, into a child by default | D4. Channel transcript passes only when both are on that channel | +| ER-10 | No child makes boards the mint door for sessions, DMs, posts or sub-agents, and no child introduces assignable-channel routing or Project-as-required | D1, D3. Conversation stays off the board | +| ER-11 | No child expands the L1 `TaskStatus` enum. Queue columns — idle, ready-for-review, waiting-on-you — are **L2 views** over existing status plus claim, assignee, park reason and seat idle | A second hold status was already refused. Seat idle is a runtime view, not a status | +| ER-12 | No child builds a second `WorkerRegistry`, a mega-loader, a mega-bus or a parallel index of truth, and none rebuilds **Graft** for membership or DM lookup | D5. One scan fills one shared map | +| ER-13 | No child closes an item on the Architect's **Still open** list without the owner | Exact package schema · reuse-vs-create · inventory helper-vs-resource · nested board cascade timing · how many POCs before a ship cut. Exploration may lean; leaning is not closing | +| ER-14 | **No W4 *ship* PR merges while W3 ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351)) has open children.** Today: FIX-1377, FIX-1416, FIX-1435, FIX-1441. Filing, specs and POCs run now | D2. The fence is on merge, not on work. Checked by the epic wake before any child's impl PR merges | + +## How the set is run + +| # | Rule | Because | +|---|---|---| +| ER-15 | A cross-cutting question a child hits is commented **up** on the epic PR, not decided locally | [DECISIONS.md](DECISIONS.md) is the single place. A local answer is a second authority — and five walls are live | +| ER-16 | A child's Linear state is mirrored the moment it changes | The epic wake derives blocked-by from Linear. **Already failing:** FIX-1377 and FIX-1416 read *In Spec Review* in Linear while both spec PRs carry the `spec approved` label and are closed | +| ER-17 | Every child's route reads *spec* by default; only a `Bug` label re-routes it | Fail-closed routing. All seven children carry `Design` + `Feature`, so all seven spec | +| ER-18 | A child depending on FIX-1377 or FIX-1416 writes against their **approved specs** and says so by name; neither has landed | FIX-1377 puts the team layer on the `WorkerConfig` contract as its own key, composed framework-default → team → seat. FIX-1416 makes the blocks scan the tool catalog behind the `tools:` fence, with a worker-colocated folder joining the seat's declared list. A child that writes as if either exists is writing against unbuilt code | +| ER-19 | **OWNER-LESS.** The project's **PR-5** — a new surface uses the settled Layer 2 name, never the superseded one — is assigned to **this epic** and to no child in it | Carried from the project spec ([#1818](https://github.com/fixpoint-labs/flow-state-dev/pull/1818)). Recorded as a gap rather than dropped ([Open 3](DECISIONS.md#open)) | + +## The proof + +| # | The epic's first cut is done when | Proved by | +|---|---|---| +| ER-20 | A piece of work filed on a **channel or org board** wakes a seat that runs it, and a coordinator seat assigns work to **team** seats — on the real path, with the seats and channels declared in files | **Unowned today.** FIX-1430 is the only candidate; adopting it is [sign-off item 2](SPEC.md#sign-off) | +| ER-21 | The queue a coordinator sees is a **view** over existing task state and seat idle, with no new status and no new L1 type | FIX-1430's goal check · ER-11 | +| ER-22 | The docs teach work routing as something a team gets by attaching a board to a channel, not something it builds — and teach `assignee` and *seat* as two different things | The epic's docs pass · ER-5 | + +**ER-20 is the rule to read twice.** The epic has a behavioural exit gate and, until the gate +ratifies FIX-1430 into the role, nothing in the set runs it. An epic that wraps on ER-1, ER-2 and +ER-3 shipping would be claiming routing works rather than having seen it. diff --git a/spec/_epics/work-routing-package-cohesion/DECISIONS.md b/spec/_epics/work-routing-package-cohesion/DECISIONS.md new file mode 100644 index 0000000000..132a0bd0c1 --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/DECISIONS.md @@ -0,0 +1,160 @@ +# FIX-1407 · Decisions + +[Spec](SPEC.md) · **Decisions** · [Rules](BUSINESS-RULES.md) · [Plan](PLAN.md) + +The calls that sit above any single issue: what was chosen, what lost, and what each locks in for +the seven issues under it. Three are the sign-off surface. All five are **Architect locks carried +into this document**, not calls taken here — FIX-1407's *FSD Architect — EM guidance* section and +each child's own fences are the source, and its **Still open** list stays open +([below](#open)). + +## The tree + +```mermaid +flowchart TD + E["FIX-1407"] -->|"kills the full cascade as the gate"| D1["D1 · the exit gate is one hop"] + E -->|"kills W4 as a W3 floor child"| D2["D2 · soft-after for ship only"] + E -->|"kills one noun for two registries"| D3["D3 · workers do, channels hold"] + E -->|"kills the ambient transcript dump"| D4["D4 · the wire and the payload"] + E -->|"kills a second index of truth"| D5["D5 · inventory is a query"] +``` + +Each edge names what the decision killed. The alternative that lost, and why, is in the card. + + +## D1 · The W4 first cut is one hop: channel/org board → seat runs, and assign team seats + +| | | +|---|---| +| **Instead of** | Making the full nested cascade — channel/org board → personal board → request board → seat — the exit gate | +| **Because** | The routing claim only needs one hop to be true or false. A cascade proves the same thing three times and cannot be finished on a schedule, which is how an epic with a behavioural gate never closes | +| **Locks in** | Personal / request nested decomposition is **phase-2 inside W4**, promoted by a POC that proves need, not by default. Three children (FIX-817, FIX-1415, and FIX-1430's nested half) sit off the gate and gate nothing. Conversation, DM and transcript stay off the board entirely — they belong to the inventory and ChannelFlow | + +**What would change my mind on the objective:** evidence that one hop is already possible today by +composition, so the set is documenting rather than building. Then W4 is a docs pass with an +inventory attached, and the shape is wrong. + + +## D2 · W4 *ship* tickets are soft-after W3; W4 is not a child of W3 + +| | | +|---|---| +| **Instead of** | Nesting W4 under W3 as a floor child · or letting a W4 ship PR merge onto a floor still being edited | +| **Because** | Every W4 child reads the file surface W3 is still closing. Nesting makes one epic that cannot wrap; ignoring the order means a package format lands against a `tools:` fence and a `TEAM.md` layer that are still specs | +| **Locks in** | Filing, specs and **parallel POCs run now** on every child. No W4 *ship* PR merges while W3 has open children — today FIX-1377, FIX-1416, FIX-1435 and FIX-1441 ([ER-14](BUSINESS-RULES.md)). The relation is soft-after / soft-related in Linear, never parent | + + +## D3 · Workers and seats *do*; channels *hold*. A board assignee key is not a Workforce seat + +| | | +|---|---| +| **Instead of** | One noun for both — merging the board's assignee registry with the Workforce roster · or an assignable channel that claims work as an executor | +| **Because** | They are different lifetimes with different identity rules. A board assignee is a key on a ledger; a seat is a roster slot that mints a flow instance. Collapsing them means every board row implies a hired seat, and every seat implies a board | +| **Locks in** | The two map **by composition**, per child. No assignable-channel routing, no channel-as-claim-worker, no second `WorkerRegistry`. `defaultWorker` is triage composition and `assignTask` is composition, not new primitives. Also locked: **Agent is an opinionated default kind**, not a Layer-1 substrate type, and there are no thin/fat seat labels | + + +## D4 · Dispatch is the wire; handoff content is the payload policy — they are not competing forks + +| | | +|---|---| +| **Instead of** | Choosing between "pass the parent session" and "pass a brief" · or a silent full parent transcript as the default | +| **Because** | One is mechanism and one is policy, and reading them as a fork is what produces both an ambient dump *and* a second runtime. A sub-agent is same-session background work with a different prompt; a worker assign is a roster seat in a **linked** session. Both need the link; only the second needs a payload rule | +| **Locks in** | Dispatch always passes `parentSessionId`, and the child binds its parent **for life at mint**. Parent history reaches a child through **opt-in tools**, never an ambient dump, and a parent's private seat history never auto-bleeds. The default payload is a required brief — `goal` / `constraints` / `acceptance` / `links` — plus channel transcript **only if both are on that channel**, and opt-in named packs. A channel session and a parent worker session may coexist with separate tool packs. **Shipped on FIX-1408**, as a decision rather than as code | + + +## D5 · Runtime inventory is an L2 *query* over the one shared scan, never a second registry + +| | | +|---|---| +| **Instead of** | A parallel mega-index of seats and channels · a second `WorkerRegistry` · rebuilding Graft for membership and DM lookup | +| **Because** | The shared Workforce map already exists and one scan already fills it. A second source of truth is a staleness bug with a schedule, and it is the thing three of this epic's children would each be tempted to build separately | +| **Locks in** | Inventory reads that map plus hire, `channelInstances` and open sessions. FIX-817's manifests read the same surface rather than building their own; FIX-1415's invite / find consumes it. `team.*` wildcards are a **later caller**, not part of the first ship, and do not block the channels convention. **The owner's stamped lean** is that ChannelFlow maintains an org-scoped resource — a lean, not a close: helper-vs-resource, eager-vs-lazy expansion, and DM as shared-kind-session vs special lane all stay on the [Still open](#open) list | + +## Who owns what + +![Who owns what: a matrix of six cross-cutting rules against the seven issues in the set. The board is the work plane is built by FIX-1385 and consumed by FIX-1430. One package format is built by FIX-1394 and consumed by FIX-1430. One runtime lookup is built by FIX-1405 and consumed by FIX-1385, FIX-817 and FIX-1415. The wire and the payload is decided by FIX-1408 and consumed by FIX-1385, FIX-1394 and FIX-1430. Assignee is not a seat is decided by FIX-1385 and consumed by FIX-1394 and FIX-1430. The last row, the propagation pass inherited from the project as PR-5, has no owner in the set and is drawn as a single gap spanning every column.](figures/ownership.svg) + +Every rule in the set has one owner, with one exception drawn as a gap: **ER-19, the project's +PR-5 propagation pass, is assigned to this epic at project altitude and to no child in it**. Read +a row to see where a decision is made, where it is built, and where it is only consumed; a +*consumes* cell is a place a child must not re-decide. FIX-1430 consumes four rows and owns none, +which is what would make it a proof. + +## Decided in review, recorded so no child reopens them + +Nothing yet — this is the epic-spec's first dispatch and the PR has had no review. Cross-cutting +answers land here as they arrive, under [ER-15](BUSINESS-RULES.md). + +## What the end-state POC showed + +**No end-state POC has been built.** The division into issues came from the Architect's child +ownership carve rather than from a sketch, and four of the seven children are themselves +POC-first — FIX-1394 explicitly so. If the gate wants the division tested before any of it starts, +an end-state POC on this branch is the instrument, and the question it would answer is whether the +package format and the board's assign surface are really two issues. + + +## Open + +**1. Does the set ratify at seven — and is FIX-1430 the epic's proof?** +*(Decides: the owner. Blocks: nothing today; it decides what the epic promises and when it can +wrap.)* + +- **The fork.** Keep all seven with two marked phase-2 and FIX-1430 adopted as the proof — or cut + the epic back to the four the body names, and let FIX-817 and FIX-1415 stand on their own? +- **In plain terms.** The epic promises one behaviour: work filed for a team reaches a seat that + runs it. Three of the seven issues build that. One (FIX-1430) is the only thing that would + *demonstrate* it, and right now the epic has a behavioural promise and nothing that checks it. + The other two — a catalog an agent can read to plan, and letting a seat create and invite to + channels — are real work that serves the same area but not the promise; both tickets fence + themselves off a ship until other things land. +- **The trade-off.** Keeping seven means one epic owns the whole routing surface, its two + phase-2 children stay next to the inventory they read, and a reader sees the real backlog. It + also means a set with three of seven on the path to its own gate — the shape where a tail grows + while the proof never starts, which is exactly what W3's necessity check was watching. Cutting + to four keeps the epic honest about what it promises, and costs two issues re-homed with no + epic holding their shared dependency on FIX-1405. +- **My recommendation: keep seven, mark FIX-817 and FIX-1415 phase-2 explicitly, and adopt + FIX-1430 as the proof.** A behavioural exit gate with no child that runs it is the defect worth + fixing here; FIX-1430 already describes exactly the gate's one hop, and its own nested-board half + is already fenced to phase-2 by its Architect section. The two phase-2 children cost nothing + while they gate nothing, and splitting them from FIX-1405 would put one manifest surface in two + epics — which is what [D5](#d5) exists to prevent. +- **What would change my mind.** If you want W4 to wrap fast on the three named issues, cut it to + four: the proof can be a fifth issue filed against the gate rather than an existing showcase + adopted into the role. And if FIX-1430 is intended as a DevForce Lab artefact rather than a + framework proof, adopting it makes the epic's finish depend on a Lab's schedule. +- **If wrong.** Two issues sit in an epic that wraps before they start, and get re-parented — a + tracker edit. Or the epic wraps on three shipped surfaces with nothing having run them, which is + the more expensive half, because "work routes to a seat" would be a claim rather than a result. + +**2. Who owns FIX-1408's unclosed walls?** *(Decides: the owner, with the Architect. Blocks: +nothing today; it decides where the reuse-vs-create policy lands.)* FIX-1408 closed as **Done** +with its *locked distinctions* recorded ([D4](#d4)) and its own open walls untouched: +reuse-vs-create session policy after POC, auto-scale / busy-copy behaviour, the +hire-or-dispatch-with-parent API naming, which opt-in history packs are v1, and how a sub-agent's +background work surfaces on a board row. Those are cross-cutting — FIX-1394's and FIX-1430's POCs +both press on them — and they now have no owning issue. **Recommendation:** let FIX-1394's POC +matrix answer reuse-vs-create and record the answer here, rather than re-opening a closed ticket. +**If wrong:** the first child to hit it decides locally, which is the second-authority failure +[ER-15](BUSINESS-RULES.md) exists to prevent. + +**3. The project's PR-5 propagation pass is assigned to this epic and to no child in it.** +*(Decides: the owner. Blocks: nothing.)* The project spec +([#1818](https://github.com/fixpoint-labs/flow-state-dev/pull/1818)) names FIX-1407 as PR-5's +owner — *a new surface uses the settled Layer 2 name, never the superseded one* — checked at "the +propagation pass". No child in the set carries that work, and the epic body does not mention it. +Recorded as [ER-19](BUSINESS-RULES.md) with no owner rather than quietly dropped. It is a gap, not +a refusal; nothing in the set is blocked by it. + +## How it got here + +- **Stood up (Sep 18)** — four documents from FIX-1407's D-4 body and its Architect EM guidance, + plus each child's own fences. No product direction was added: the exit gate, the sequencing, the + vocabulary and every invent-kill are carried, and the Architect's **Still open** list is carried + open. +- **The set was written as seven, not four (Sep 18)** — FIX-817, FIX-1415 and FIX-1430 are + parented in Linear but absent from the epic body. Recorded in the set table with the split + visible, and put to the gate as Open 1 rather than resolved on the branch. +- **FIX-1408 recorded as done-by-decision (Sep 18)** — no implementation PR; its locked + distinctions became D4 and ER-4, and its unclosed walls became Open 2. diff --git a/spec/_epics/work-routing-package-cohesion/PLAN.md b/spec/_epics/work-routing-package-cohesion/PLAN.md new file mode 100644 index 0000000000..4be348e5d2 --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/PLAN.md @@ -0,0 +1,101 @@ +# FIX-1407 · Plan + +[Spec](SPEC.md) · [Decisions](DECISIONS.md) · [Rules](BUSINESS-RULES.md) · **Plan** + +An epic plan sequences the work and says what each piece entails. It does not say how to build any +piece; that's each issue's own plan. IDs cross-reference [DECISIONS.md](DECISIONS.md) (D-n) and +[BUSINESS-RULES.md](BUSINESS-RULES.md) (ER-n). + +## The path + +![The path: a ship fence band across the top marking W3 open with four children, then two W3 input lanes for TEAM.md and the tools fence whose specs are approved and unbuilt, then seven issue lanes against time. FIX-1408 has a done bar behind the now line at September 18. The three first-cut lanes — package cohesion, runtime inventory and channel boards — are empty ahead of the now line and marked spec and POC may start now, ship merge held by the fence. The two phase-2 lanes for catalog manifests and channel-admin sit further right, off the gate. The manager-queue lab lane is last and marked proof, unadopted. The critical path is drawn in the gutter through the inventory, the boards and the lab.](figures/path.svg) + +Nine lanes: two inputs from W3 and the seven children. **One bar exists** — FIX-1408, done by +decision — and everything else is an empty lane, because nothing else has started. That is the +honest picture at the gate, and it is why the band across the top matters more than any lane: the +fence is on **merge**, not on work ([ER-14](BUSINESS-RULES.md)), so all six remaining children can +open a spec and a POC the day the objective is approved. The critical path is short and runs +inventory → boards → the proof; package cohesion is the one lane that can run beside it the whole +way. The dependency graph itself is in +[the spec](SPEC.md#how-the-issues-flow-into-each-other); this document adds time to it. + +## What each issue entails + +| Issue | Route | Consumes | Delivers | Releases | Size | +|---|---|---|---|---|---| +| **FIX-1408** dispatch policy | design → **closed as decided** | D4's two leans | Sub-agent vs seat-assign, `parentSessionId` at mint, opt-in history, the brief | Every other child | — | +| **FIX-1405** runtime inventory | spec → POC → impl | The shared scan · hire · `channelInstances` · open sessions (D5) | One L2 lookup for membership expand, author check, fan-out, DM find-or-create | FIX-1385 · FIX-817 · FIX-1415 · `team.*` wildcards, later | Medium | +| **FIX-1385** channel boards | spec → POC → impl | The ChannelFlow floor · the inventory (ER-3) · ER-5 | A channel holding `0..N` TaskCollections; two doors, one mutation surface | The proof · FIX-1430 | Large | +| **FIX-1394** package cohesion | **POC matrix → ratify** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both unlanded** (ER-18) · D4 | One package format, two attachment modes; the scale-back before both surfaces grow twice | Ship tickets, cut after ratify | Large | +| **FIX-817** catalog manifests · *phase-2* | spec → impl | The inventory as a reader, not a second index (D5) | One discovery shape across seats, channels, resources, tools, skills | Planning before assign | Medium | +| **FIX-1415** channel-admin · *phase-2* | spec → ratify → ship tickets | The inventory · the `tools:` fence · Collab's system-vs-dynamic lanes | Channel create / delete / invite as a **capability**, behind seat `tools:` | DevForce's "EM opens a channel" pressure | Medium | +| **FIX-1430** manager-queue lab · *proof, unadopted* | spec → goal check | ER-1 · ER-2 · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, with queue columns as views | The epic's wrap | Medium | + +**Size is a read, not an estimate.** Four of the seven are exploration tickets whose own scope is +the thing being explored; FIX-1394 is the clearest case — its deliverable is a POC matrix, and the +ship tickets it cuts are not in this set yet. + +## Where it is + +Status lives in one place: [the set table in the spec](SPEC.md#the-set--as-of-2026-09-18). The +lanes above carry the same state as a picture of time and are redrawn when it moves. + +**The inputs from other epics, and their verified state.** FIX-1377 (optional `TEAM.md`) and +FIX-1416 (custom tools authoring) are W3 children whose specs are **approved** — +[#1819](https://github.com/fixpoint-labs/flow-state-dev/pull/1819) and +[#1890](https://github.com/fixpoint-labs/flow-state-dev/pull/1890), both carrying the +`spec approved` label, both closed unmerged — and whose code has **not landed**. Linear still reads +*In Spec Review* for both, which is [ER-16](BUSINESS-RULES.md) already failing on the two +dependencies this epic cares most about. Treat the approved specs as the contract (ER-18); treat +the Linear state as stale. + +## What unblocks what, from here + +1. **The objective is approved** → all six live children may open a spec and a POC immediately. + Nothing in the set waits on anything else to *start*; the fence is on merge (ER-14). +2. **FIX-1405's inventory lands** → FIX-1385 can assign a **team** seat rather than a literal id, + and FIX-817 and FIX-1415 stop being blocked on a lookup that doesn't exist. This is the one + edge that genuinely sequences. +3. **FIX-1385's board surface lands** → the proof can run. Until then ER-20 has nothing behind it. +4. **W3's last child merges** → the ship fence lifts and W4 impl PRs may merge. Four remain: + FIX-1377, FIX-1416, FIX-1435, FIX-1441. +5. **FIX-1394's POC matrix is ratified** → and only then are its ship tickets cut (ER-8). Those + tickets are new children of this epic and will need the set table and this path redrawn. +6. **If the gate adopts FIX-1430 as the proof** → ER-20 gets an owner and the epic gains a + finishing condition. If it doesn't, the epic needs one filed, or it wraps on a claim. + +## Coordination seams to watch + +| Seam | Between | Rule | +|---|---|---| +| The inventory surface | FIX-1405 and FIX-817 | 1405 owns it; 817 reads it. Whether the inventory resource **is** the manifest or only feeds a tool is an open wall — neither child answers it alone (D5, ER-13) | +| The seat's `tools:` fence | FIX-1394, FIX-1415, and FIX-1416 outside the epic | Three edits pressing on one fence. Neither W4 child widens past seat `tools:`, and both write against FIX-1416's approved spec rather than around it (ER-18) | +| The seat's prompt compose | FIX-1394 and FIX-1377 outside the epic | The package format composes onto framework default → team → seat. It does not invent a fourth layer, and it does not merge the team's text into the seat's | +| The board's claim / assign surface | FIX-1385 and FIX-1430 | 1385 owns it; the lab consumes it. Cross-flow seats must declare the **same logical board** (`boardId` + ledger + own same-flow gating dispatcher) — whether that is a file-convention teach is FIX-1385's open wall, and the lab does not answer it | +| `goals/devforce-lab/` | FIX-1430 and FIX-1426 (Done, outside the set) | 1426 is the coding seam only and does not grow into the manager-queue showcase | +| The Workforce docs section | FIX-1385, FIX-1405, FIX-1430 | All three teach boards. Whichever lands second links rather than repeats, and none teaches ahead of its reader | + +## Not children, deliberately + +[FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351) (**W3**) — soft-after for ship, +soft-related otherwise, never re-parented (D2). +[FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341) (**Collab RC**) — parked; FIX-1415 +shapes a capability, not a reopen. +[FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359) (**OOTB agent kind**) — shipped, +parallel history, not reopened inside W4. +Also consumed, not owned: [FIX-1388](https://linear.app/fixpoint-labs/issue/FIX-1388) (Door B), +[FIX-1393](https://linear.app/fixpoint-labs/issue/FIX-1393) (the `tools:` fence), +[FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426) (DevForce first slice, Done), +[FIX-1440](https://linear.app/fixpoint-labs/issue/FIX-1440) (nested-session removal — cited, not a +ship child), [FIX-1410](https://linear.app/fixpoint-labs/issue/FIX-1410) (D-12 DevForce Lab), +[FIX-1373](https://linear.app/fixpoint-labs/issue/FIX-1373) (board-worker vocab park) and +[FIX-957](https://linear.app/fixpoint-labs/issue/FIX-957) (durable task collections). + +## Wrap + +When ER-20 holds: run the lessons pass over the set's review rounds; dispatch the docs polish over +the Workforce pages the children each edited in isolation — it owes the sentence nobody else will +write, that a board **assignee** and a Workforce **seat** are two different things that map by +composition (ER-5, ER-22). Then settle whether PR-5's propagation pass ran or is handed on +([ER-19](BUSINESS-RULES.md)), refresh the set table and the path one last time, and close the epic +PR unmerged. diff --git a/spec/_epics/work-routing-package-cohesion/SPEC.md b/spec/_epics/work-routing-package-cohesion/SPEC.md new file mode 100644 index 0000000000..086acd005e --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/SPEC.md @@ -0,0 +1,128 @@ +# FIX-1407 · W4: work reaches a seat + +**Spec** · [Decisions](DECISIONS.md) · [Rules](BUSINESS-RULES.md) · [Plan](PLAN.md) + +Epic · 7 issues · Workforce: Layer 2 Abstraction · Goal 1 — Workforce multi-seat real usage / +architecture cohesion + +## Four teams, before and after + +| A team that… | Today | After this epic's first cut | +|---|---|---| +| **has a piece of work for a team of seats** | Nowhere shared to put it. Work is dispatched by hand, from code, one seat at a time | It goes on the channel's board. A seat claims it or is assigned it, and runs it | +| **writes one package for a seat** | Two overlapping systems — the seat's own prompt file and skills — and no answer to which one a tool belongs in | One package format, two attachment modes: always-on for a seat, opt-in for a library | +| **asks what seats and channels exist right now** | Opaque strings. `engineering.*` cannot expand, and a seat asking another seat has no way to find their DM | One runtime lookup answers membership, author checks, fan-out and DM find-or-create | +| **hands work from one seat to another** | Invents an ambient transcript dump, or a second runtime | A named brief, a linked session bound to its parent for life, and parent history only through opt-in tools | + +**Why now.** W3 made a Workforce *describable* — workers, channels, resources and skills all +declare in one tree. Nothing yet says how **work reaches** one of those declared seats. Four +surfaces are missing or disagree: the package story (a seat's prompt file versus skills), what +exists at runtime (no inventory, so membership is an opaque string), what *assign* means (a +sub-agent, a roster seat, a board row — all three, unreconciled), and where work lives at all. +That gap is why a declared team is still a demo. This epic closes the routing half of Goal 1: W3 +proved a seat is *described*, W4 proves a seat is *given work*. It does not move the corpus-wide +goal count on its own — the lead measure moves when a child ships a goal check. + +## What's in the box + +![What's in the box: the W4 first cut is a channel or org board that routes work to a seat run and assigns team seats — the exit gate, one hop — plus one package format with two attachment modes, a runtime inventory of seats and channels as one L2 query, and the shipped dispatch and session policy. Composed in by the app: custom kinds as flow factories, and defaultWorker triage. A phase-2 panel inside W4, off the exit gate, holds the nested personal and request board cascade, channel-admin verbs, catalog manifests and the manager-queue lab's nested half. Below a fence, what is not built: no Agent, Channel, Team, MessageBoard, TeamFlow or SessionBoard L1 type, no Collab RC, no silent parent transcript, no merged board-assignee and seat registry, no team wildcards as first ship, no second WorkerRegistry or mega-loader, no assignable-channel routing, no BoardFlow as the only mint, no Graft rebuild.](figures/end-state.svg) + +Everything inside the box is what a team gets once the first cut lands. The **exit gate is the top +row and only the top row**: a channel or org board routes work to a seat run, and can assign team +seats ([D1](DECISIONS.md#d1)). The strip under it is phase-2 *inside* W4 — named so nobody reads +it as refused, and fenced off the gate so the epic can finish. The bottom strip is what the set +refuses to build, and every item in it was named out by the Architect rather than forgotten. + +## The set · as of 2026-09-18 + +This table is the live one. It is refreshed on the epic PR as issues move; the plan and the +figures point here rather than repeating it. + +**The epic body names four.** Three more were parented after it was written, and whether they +belong is the one thing the objective gate has to settle — see [sign-off](#sign-off) and +[Open 1](DECISIONS.md#open). + +| Issue | What it delivers | Why the set needs it | Status | +|---|---|---|---| +| FIX-1408 | Dispatch / session policy — sub-agent is same-session background, worker assign is a roster seat in a **linked** session, `parentSessionId` always at mint, history by opt-in tools | The wire everything else routes over. Locked as decision, not code | **Done** · closed 2026-09-18 · no impl PR ([below](#a-note-on-fix-1408)) | +| FIX-1394 | Package cohesion — one package format, two attachment modes (seat always-on, library opt-in). POC-first | Half the epic's title. Two overlapping surfaces grow board and tool sugar twice until this settles | Not started · Backlog | +| FIX-1405 | Runtime inventory of seats + channels; lean is an org-scoped resource ChannelFlow maintains | Assigning a **team** seat needs to know which seats exist. Without it the board can route to a literal id and nothing else | Not started · Backlog | +| FIX-1385 | Channel workstreams + boards — a channel holds `0..N` TaskCollections; channel actions and `taskTools` are two doors on one surface | **The exit gate's surface.** It is the board in "channel/org board → seat runs" | Not started · Todo | +| FIX-817 | Collection / catalog manifests + agent introspection — one discovery shape over seats, channels, resources, tools, skills | Planning *before* assign. Not on the path to the exit gate; it reads the same inventory FIX-1405 builds | Not started · Todo · `Open Question` | +| FIX-1415 | Channel-admin capability — create / delete / invite behind the seat's `tools:` fence | Design only until Collab park and inventory land, by its own Architect fence. Off the exit gate | Not started · Todo | +| FIX-1430 | Manager-queue lab — a coordinator seat owns a channel + workstream board, assigns to linked seats, and shows queue state as **views** over existing task status | The only child shaped like a **proof** of the exit gate. Nothing else in the set runs it end to end | Not started · Todo | + +1 done · 0 in flight · 6 not started. Four are the epic body's set; three (FIX-817, FIX-1415, +FIX-1430) were parented later and the body does not name them. + +**Is seven really four?** Seven is **three plus a proof plus two phase-2** — and the counting is +the gate's call, not mine. FIX-1394, FIX-1405 and FIX-1385 are the first cut the body names. +FIX-1430 is the only child that would *run* the exit gate, and a set with a behavioural gate and +nothing that runs it is missing its proof, so my recommendation is to adopt it as one. FIX-817 and +FIX-1415 both sit behind FIX-1405 and both fence themselves off a ship; kept as phase-2, they gate +nothing and keep one shared surface with its owner. The full trade, including the arm where two of +them leave the epic, is [Open 1](DECISIONS.md#open). + + +**A note on FIX-1408.** It went Backlog → Done on 2026-09-18 with no implementation PR: what +shipped is the **decision**, recorded as [D4](DECISIONS.md#d4) and [ER-4](BUSINESS-RULES.md). Its +own open walls are not closed by that and now have no owning issue +([Open 2](DECISIONS.md#open)). + +## How the issues flow into each other + +```mermaid +flowchart LR + T377["FIX-1377 · TEAM.md · W3, unlanded"] -.->|"the team layer a package composes onto"| PK + T416["FIX-1416 · tools fence + blocks scan · W3, unlanded"] -.->|"what a seat may call"| PK + DP["FIX-1408 · dispatch / session policy"] -->|"the wire: parentSessionId, opt-in history"| BD + DP --> PK["FIX-1394 · package cohesion"] + INV["FIX-1405 · runtime inventory"] -->|"which seats exist to assign to"| BD["FIX-1385 · channel boards"] + BD -->|"the exit gate's surface"| LAB["FIX-1430 · manager-queue lab"] + PK -->|"one package a seat carries"| LAB + INV -->|"the same manifest shape"| CAT["FIX-817 · catalog manifests · phase-2"] + INV -.->|"invite and find"| ADM["FIX-1415 · channel-admin · phase-2"] + T416 -.-> ADM + classDef done stroke-width:2px + class DP done +``` + +A solid edge is what one issue hands the next and blocks until it lands. **Dashed edges from the +left are inputs from W3 that have not landed** — FIX-1377 and FIX-1416 are spec-approved and still +to be built, so no W4 child may write as though they exist ([ER-18](BUSINESS-RULES.md)). The +dashed edge into FIX-1415 is soft: its own fence holds it for Collab and inventory, so it waits +rather than blocks. A heavy border is done. Every node is filed, so the set carries no +placeholders. + +## What stays as it is + +- **Collab RC ([FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341)).** Dynamic rooms, + membership mutate and the roster stay parked. FIX-1415 shapes a capability surface; it does not + reopen Collab. +- **The W3 floor ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351)).** A sibling epic, + soft-*after* for ship and never a parent ([D2](DECISIONS.md#d2)). +- **The OOTB agent kind ([FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359)).** Shipped, + parallel history. Package cohesion flags drift against it; it is not redesigned from zero. +- **The board's claim system.** One task board, one claim system. `assignee` stays optional and + `TaskStatus` does not grow ([ER-11](BUSINESS-RULES.md)). +- **`goals/devforce-lab/` ([FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426), Done).** + The coding seam only. It does not grow into FIX-1430's showcase. + +## Sign off + +1. **[D1](DECISIONS.md#d1) · The exit gate is channel/org board → seat runs / assign team seats. + The nested personal/request cascade is phase-2.** If wrong: the epic either finishes on a cut + too thin to prove routing, or never finishes because the gate kept moving. +2. **[Open 1](DECISIONS.md#open) · Does the set ratify at seven, and is FIX-1430 the proof?** + Three issues were parented after the body was written. Two of them (FIX-817, FIX-1415) are + phase-2 by their own fences; the third (FIX-1430) is the only child shaped to run the gate, and + the epic currently has a behavioural exit gate with **nothing that proves it**. The full + six-part ask is in [DECISIONS.md → Open](DECISIONS.md#open). +3. **[D2](DECISIONS.md#d2) · W4 *ship* PRs are soft-after W3; filing, specs and POCs run now.** If + wrong: either W4 ships onto a floor that moves under it, or four children idle for a floor that + was never going to block them. + +**Open: two.** Whether the set is seven, and who owns FIX-1408's unclosed walls +([DECISIONS.md → Open](DECISIONS.md#open)). The rules every child obeys are in +[BUSINESS-RULES.md](BUSINESS-RULES.md); the order the work runs in, and what each issue entails, +is [PLAN.md](PLAN.md). diff --git a/spec/_epics/work-routing-package-cohesion/figures/end-state.svg b/spec/_epics/work-routing-package-cohesion/figures/end-state.svg new file mode 100644 index 0000000000..12a1a55646 --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/figures/end-state.svg @@ -0,0 +1,78 @@ + + + + + + IN THE BOX — THE W4 FIRST CUT + + + THE EXIT GATE + channel / org board → seat runs · assign team seats + + + one package format + two attachment modes: seat always-on · library opt-in — after POC ratify + + + runtime inventory of seats + channels + one L2 query: membership · author check · fan-out · DM find-or-create + + + dispatch + session policy + parentSessionId at mint · history by opt-in tools — shipped, FIX-1408 + + A sub-agent is same-session background; a worker assign is a seat in a linked session. + + + + + COMPOSED IN BY THE APP + + custom kinds + flow factories under workforce/flows/ + + defaultWorker triage + who to assignTask to — composition + + + + + PHASE-2 INSIDE W4 — OFF THE GATE + + nested personal / request boards + channel-admin verbs · catalog manifests + the manager-queue lab's nested half + + + + + NOT BUILT + no Agent / Channel / Team / MessageBoard / TeamFlow / SessionBoard as an L1 type · no Collab RC as this epic + no silent parent transcript · no merged board-assignee + seat registry · no team.* wildcards as first ship + no second WorkerRegistry or mega-loader · no assignable-channel · no BoardFlow-as-only-mint · no Graft rebuild + + diff --git a/spec/_epics/work-routing-package-cohesion/figures/ownership.svg b/spec/_epics/work-routing-package-cohesion/figures/ownership.svg new file mode 100644 index 0000000000..7cc773156b --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/figures/ownership.svg @@ -0,0 +1,107 @@ + + + + WHO OWNS WHAT — EVERY RULE HAS ONE OWNER, EXCEPT THE LAST + + + FIX-1385 + FIX-1394 + FIX-1405 + FIX-1408 + FIX-817 + FIX-1415 + FIX-1430 + + + + ER-1 · the board is the work plane + BUILDS + + + + + + consumes + + + + ER-2 · one package format + + BUILDS + + + + + consumes + + + + ER-3 · one runtime lookup + consumes + + BUILDS + + consumes + consumes + + + + + ER-4 · the wire and the payload + consumes + consumes + + DECIDES + + + consumes + + + + ER-5 · assignee is not a seat + DECIDES + consumes + + + + + consumes + + + + ER-19 · the propagation pass + + no owner in the set — inherited from the project as PR-5, assigned to the epic itself + + + + + the one owner — decides or builds it + + consumes it, and must not re-decide it + + no owner — a gap, not a refusal + FIX-1430 consumes four rows and owns none, which is what would make it the proof + + diff --git a/spec/_epics/work-routing-package-cohesion/figures/path.svg b/spec/_epics/work-routing-package-cohesion/figures/path.svg new file mode 100644 index 0000000000..324759eb3c --- /dev/null +++ b/spec/_epics/work-routing-package-cohesion/figures/path.svg @@ -0,0 +1,103 @@ + + + + THE PATH — ONE LANE PER ISSUE, AGAINST TIME + + + + Sep 16 + now · Sep 18 + ahead · nothing scheduled + + + + SHIP FENCE + + W3 floor open — FIX-1377 · FIX-1416 · FIX-1435 · FIX-1441 — no W4 ship PR merges until they close + + + + + + + + + + + + + FIX-1377 TEAM.md · W3 + + spec approved · unbuilt · #1819 + + FIX-1416 tools fence · W3 + + spec approved · unbuilt · #1890 + + + + FIX-1408 dispatch policy + + done by decision — no impl PR + + FIX-1405 inventory + + spec + POC may start now + + FIX-1385 channel boards + + spec + POC now — the exit gate's surface + + FIX-1394 package cohesion + + POC matrix first · ship tickets after ratify + + FIX-817 catalog manifests + + phase-2 · off the gate · waits on inventory + + FIX-1415 channel-admin + + phase-2 · off the gate · design only + + FIX-1430 manager-queue lab + + the proof — unadopted. The epic ends here + + + + + done + + an input from W3 — spec approved, unbuilt + + not started — a lane, not a bar + + + on the critical path to the proof + + From 8f8daaec68fa512349a4f58e18832c13cf9060f4 Mon Sep 17 00:00:00 2001 From: Claude Date: Fri, 18 Sep 2026 21:21:29 +0000 Subject: [PATCH 02/21] spec(epic): fold round-1 review into the W4 epic-spec (FIX-1407) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Five reviewers on #1905. What changed: - D5 becomes two layers, not a fork: a declared roster composed at read time from the three existing readers (promoting both labs' duplicated private LabRoster), and the live org resource ChannelFlow updates. ER-3 names both, ER-12 adds "compose is not a registry", ER-13 drops inventory helper-vs-resource from Still-open — closed by its own owner. - The W3 ship fence had three different release conditions across PLAN, ER-14, D2 and the path figure. Now one, everywhere: every W3 child that carries an implementation is merged to main; children completed by decision, duplicated or cancelled do not hold it. - FIX-1394 -> FIX-1430 is redrawn non-blocking. The proof consumes package cohesion's ratified contract, not its shipped package; binding it to the ship would hang the exit gate on tickets that are cut only after a POC ratify. - The set question is re-framed on the real trade-off: a non-terminal phase-2 child holds the epic's wrap open, so the exit gate and the wrap are different moments. "Phase-2 gates nothing" was false and is corrected. - FIX-1408's open walls are owned (FIX-1394's POC holds the evidence, the epic holds the decision) and FIX-1385 owns ER-19. Both leave the Open list; one question remains. - ER-23 added: a shared contract's spec is approved before the next child specs against it. ER-16's staleness example is now historical — both W3 inputs moved to In Development. - Editorial pass: 6,290 -> 5,577 words while folding all of the above in. Figures redrawn and checked in both themes; every intra-set link and rule id verified to resolve. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01PxDE76gD3BxXnPPjT7BT1T --- .../BUSINESS-RULES.md | 68 ++++--- .../DECISIONS.md | 184 +++++++++--------- .../work-routing-package-cohesion/PLAN.md | 128 ++++++------ .../work-routing-package-cohesion/SPEC.md | 160 ++++++++------- .../figures/end-state.svg | 15 +- .../figures/ownership.svg | 21 +- .../figures/path.svg | 12 +- 7 files changed, 288 insertions(+), 300 deletions(-) diff --git a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md index db0c8054a9..ead24ce2b6 100644 --- a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md +++ b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md @@ -2,61 +2,59 @@ [Spec](SPEC.md) · [Decisions](DECISIONS.md) · **Rules** · [Plan](PLAN.md) -At epic altitude the rules aren't behaviours of one feature; they're the constraints every child -spec and implementation must satisfy, and the place a cross-spec review checks. Each says who owns -it and where it's checked. ER-1 to ER-5 and ER-19 are the rows of the ownership matrix in -[DECISIONS.md](DECISIONS.md#who-owns-what). +Not behaviours of one feature — the constraints every child spec and implementation must satisfy, +and what a cross-spec review checks. Numbers are stable ids, not an order. ER-1 to ER-5 and ER-19 +are the rows of the ownership matrix in [DECISIONS.md](DECISIONS.md#who-owns-what). ## What a team gets, and what it doesn't | # | Rule | Owner | Checked at | |---|---|---|---| -| ER-1 | Work for a team lives as a row on a **channel or org board** — `0..N` TaskCollections attached to a channel — and a seat claims it or is assigned it. Channel flow actions and `taskTools` are **two doors on one mutation surface** | FIX-1385 | FIX-1385's spec · the proof | -| ER-2 | A seat carries **one package format**, with two attachment modes: seat **always-on**, library **opt-in**. A tool, a skill and an instruction have one home between them, not two | FIX-1394, after POC ratify | FIX-1394's POC matrix · every child's spec review | -| ER-3 | Membership expand, author check, fan-out and DM find-or-create are answered by **one L2 query** over the shared Workforce map plus hire, `channelInstances` and open sessions | FIX-1405 | FIX-1405's two prove-pressure callers · FIX-817's and FIX-1415's spec review | -| ER-4 | Dispatch always passes `parentSessionId` and the child binds its parent **for life at mint**. Parent history reaches a child only through **opt-in tools**; the default payload is the brief — `goal` / `constraints` / `acceptance` / `links` | FIX-1408 (D4) — **shipped as a decision** | Every child's spec review | +| ER-1 | Work for a team lives as a row on a **channel or org board** — `0..N` TaskCollections on a channel — and a seat claims it or is assigned it. Channel actions and `taskTools` are **two doors on one mutation surface** | FIX-1385 | FIX-1385's spec · the proof | +| ER-2 | A seat carries **one package format**, two attachment modes: seat **always-on**, library **opt-in**. A tool, a skill and an instruction have one home between them | FIX-1394, after POC ratify | FIX-1394's POC matrix · the proof, which may not invent a third shape | +| ER-3 | Inventory is **two layers**: a **declared roster** composed at read time from the existing readers, and the **live org resource** ChannelFlow updates as seats and channels open. Membership, author check, fan-out and DM find-or-create are answered by the live layer; the declared layer answers what the tree says exists | FIX-1405 | FIX-1405's two prove-pressure callers · FIX-817's and FIX-1415's spec review | +| ER-4 | Dispatch always passes `parentSessionId`; the child binds its parent **for life at mint**. Parent history reaches a child only by **opt-in tools**; the default payload is the brief — `goal` / `constraints` / `acceptance` / `links`. **Linked is not nested**: no child reads this as reviving the substrate FIX-1440 removed | FIX-1408 (D4) — **shipped as a decision** | Every child's spec review | | ER-5 | A board `assignee` is a **board-worker key**, not a Workforce seat; the two map by composition. `assignee` stays optional, and a registry board's "must be assigned" is discipline, not schema | FIX-1385 | FIX-1385's spec · FIX-1430's wiring | -**On ER-2**: *after POC ratify* is the rule, not a hedge. FIX-1394 does not cut a ship ticket from -exploration alone, so no other child may write against a package schema before that ratify lands — -the schema itself is on the [Still open](DECISIONS.md#open) list. - -**On ER-4**: it is discharged by a decision rather than by a check, because FIX-1408 closed -without an implementation PR. The first child to build over the wire is where it first gets tested, -and its unclosed walls are [Open 2](DECISIONS.md#open). +ER-2's *after POC ratify* is the rule, not a hedge — no child writes against a package schema +before it. ER-3's declared layer does not substitute for the live one, its public export is on the +owner's objective gate, and FIX-1405 implementation waits for that gate and its own spec. ER-4 is +discharged by a decision rather than a check, so the first child to build over the wire is where it +first gets tested. ## What no child may do | # | Rule | Because | |---|---|---| -| ER-6 | No child invents **Agent, Channel, Team, MessageBoard, TeamFlow, ChannelAdmin or SessionBoard as an L1 package type**, and none revives `materializeAgent` or an `AgentRegistry` | D3. Agent is an opinionated default *kind*; the rest are L2 opinion or nothing | -| ER-7 | No child ships `team.*` (or other) wildcards before the inventory lands, and none blocks the channels convention on them | D5. Wildcards are a later **caller** of inventory, not part of the first ship | -| ER-8 | No child ships the package collapse from FIX-1394 alone, without a POC ratify | The whole delivery posture of that ticket: POC matrix → Architect/PM ratify → then cut ship tickets | -| ER-9 | No child dumps a parent's full transcript, or a parent's private seat history, into a child by default | D4. Channel transcript passes only when both are on that channel | -| ER-10 | No child makes boards the mint door for sessions, DMs, posts or sub-agents, and no child introduces assignable-channel routing or Project-as-required | D1, D3. Conversation stays off the board | -| ER-11 | No child expands the L1 `TaskStatus` enum. Queue columns — idle, ready-for-review, waiting-on-you — are **L2 views** over existing status plus claim, assignee, park reason and seat idle | A second hold status was already refused. Seat idle is a runtime view, not a status | -| ER-12 | No child builds a second `WorkerRegistry`, a mega-loader, a mega-bus or a parallel index of truth, and none rebuilds **Graft** for membership or DM lookup | D5. One scan fills one shared map | -| ER-13 | No child closes an item on the Architect's **Still open** list without the owner | Exact package schema · reuse-vs-create · inventory helper-vs-resource · nested board cascade timing · how many POCs before a ship cut. Exploration may lean; leaning is not closing | -| ER-14 | **No W4 *ship* PR merges while W3 ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351)) has open children.** Today: FIX-1377, FIX-1416, FIX-1435, FIX-1441. Filing, specs and POCs run now | D2. The fence is on merge, not on work. Checked by the epic wake before any child's impl PR merges | +| ER-6 | No **Agent, Channel, Team, MessageBoard, TeamFlow, ChannelAdmin or SessionBoard as an L1 package type**; no revived `materializeAgent` or `AgentRegistry` | D3. Agent is an opinionated default *kind*; the rest are L2 opinion or nothing | +| ER-7 | No `team.*` (or other) wildcards before the inventory lands, and none blocking the channels convention | D5. Wildcards call the live layer later; they are not first ship | +| ER-8 | No shipping the package collapse from FIX-1394's exploration alone, without a ratify | That ticket's posture: POC matrix → ratify → then cut ship tickets | +| ER-9 | No dumping a parent's full transcript, or its private seat history, into a child by default | D4. Channel transcript passes only when both are on that channel | +| ER-10 | Boards are not the mint door for sessions, DMs, posts or sub-agents. No assignable-channel routing, no Project-as-required | D1, D3. Conversation stays off the board | +| ER-11 | No expanding the L1 `TaskStatus` enum. Queue columns — idle, ready-for-review, waiting-on-you — are **L2 views** over status plus claim, assignee, park reason and seat idle | A second hold status was already refused. Seat idle is a view, not a status | +| ER-12 | No second `WorkerRegistry`, mega-loader, mega-bus or parallel index of truth; no Graft rebuild for membership or DM lookup. **Compose is not a registry** — ER-3's declared layer derives per read, registers nothing, and may not stand in for the live resource | D5. One derivation over the readers that exist; one live resource for what is open | +| ER-13 | No closing an item on the Architect's **Still open** list without the owner: exact package schema · reuse-vs-create · nested cascade timing · how many POCs before a ship cut. Exploration may lean; leaning is not closing | *Inventory helper-vs-resource left this list in round 1* — closed as **both layers** ([D5](DECISIONS.md#d5)), by the owner himself | +| ER-14 | **No W4 *ship* PR merges until every W3 ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351)) child that carries an implementation is merged to main.** Children completed by decision, duplicated or cancelled do not hold the fence | D2. The risk is W4 shipping onto unlanded W3 *code*, so the condition is merge, not closure — this set already contains a child completed by decision, and closed-but-unmerged spec PRs are normal here. The fence is on merge, not on work | ## How the set is run | # | Rule | Because | |---|---|---| -| ER-15 | A cross-cutting question a child hits is commented **up** on the epic PR, not decided locally | [DECISIONS.md](DECISIONS.md) is the single place. A local answer is a second authority — and five walls are live | -| ER-16 | A child's Linear state is mirrored the moment it changes | The epic wake derives blocked-by from Linear. **Already failing:** FIX-1377 and FIX-1416 read *In Spec Review* in Linear while both spec PRs carry the `spec approved` label and are closed | -| ER-17 | Every child's route reads *spec* by default; only a `Bug` label re-routes it | Fail-closed routing. All seven children carry `Design` + `Feature`, so all seven spec | -| ER-18 | A child depending on FIX-1377 or FIX-1416 writes against their **approved specs** and says so by name; neither has landed | FIX-1377 puts the team layer on the `WorkerConfig` contract as its own key, composed framework-default → team → seat. FIX-1416 makes the blocks scan the tool catalog behind the `tools:` fence, with a worker-colocated folder joining the seat's declared list. A child that writes as if either exists is writing against unbuilt code | -| ER-19 | **OWNER-LESS.** The project's **PR-5** — a new surface uses the settled Layer 2 name, never the superseded one — is assigned to **this epic** and to no child in it | Carried from the project spec ([#1818](https://github.com/fixpoint-labs/flow-state-dev/pull/1818)). Recorded as a gap rather than dropped ([Open 3](DECISIONS.md#open)) | +| ER-15 | A cross-cutting question is commented **up** on the epic PR, never decided locally — including FIX-1408's unclosed walls, where **FIX-1394's POC owns the evidence and this epic owns the decision** | [DECISIONS.md](DECISIONS.md) is the single place. A local answer is a second authority | +| ER-16 | A child's Linear state is mirrored the moment it changes | The epic wake derives blocked-by from Linear. Caught once: FIX-1377 and FIX-1416 read *In Spec Review* while both spec PRs were closed and approved; corrected to In Development on 2026-09-18 | +| ER-17 | Every child routes to *spec* by default; only a `Bug` label re-routes | Fail-closed. All seven carry `Design` + `Feature`, so all seven spec | +| ER-18 | A child depending on FIX-1377 or FIX-1416 writes against their **approved specs** and names them; neither has landed | FIX-1377 puts the team layer on `WorkerConfig` as its own key, composed framework-default → team → seat. FIX-1416 makes the blocks scan the tool catalog behind the `tools:` fence, a worker-colocated folder joining the seat's declared list | +| ER-19 | The project's **PR-5** — a new surface uses the settled Layer 2 name, never the superseded one — is owned by **FIX-1385** | From the project spec ([#1818](https://github.com/fixpoint-labs/flow-state-dev/pull/1818)). FIX-1385 mints the largest new vocabulary surface in the set | +| ER-23 | **A shared contract's spec is approved before the next child specs against it.** FIX-1405's reader contract closes before FIX-817 starts; the `tools:` fence settles across FIX-1416 → FIX-1394 → FIX-1415 in that order | Two children designing one discovery API at the same time is how a set ships two of them. Orders the spec work without re-parenting | ## The proof | # | The epic's first cut is done when | Proved by | |---|---|---| -| ER-20 | A piece of work filed on a **channel or org board** wakes a seat that runs it, and a coordinator seat assigns work to **team** seats — on the real path, with the seats and channels declared in files | **Unowned today.** FIX-1430 is the only candidate; adopting it is [sign-off item 2](SPEC.md#sign-off) | -| ER-21 | The queue a coordinator sees is a **view** over existing task state and seat idle, with no new status and no new L1 type | FIX-1430's goal check · ER-11 | -| ER-22 | The docs teach work routing as something a team gets by attaching a board to a channel, not something it builds — and teach `assignee` and *seat* as two different things | The epic's docs pass · ER-5 | +| ER-20 | Work filed on a **channel or org board** wakes a seat that runs it, and a coordinator seat assigns work to **team** seats — on the real path, seats and channels declared in files | **Unowned today.** FIX-1430 is the only candidate; adopting it is part of the set question ([Open](DECISIONS.md#open)) | +| ER-21 | The queue a coordinator sees is a **view** over existing task state and seat idle — no new status, no new L1 type | FIX-1430's goal check · ER-11 | +| ER-22 | The docs teach routing as something a team gets by attaching a board to a channel, not something it builds — and teach `assignee` and *seat* as two different things | The epic's docs pass · ER-5 | -**ER-20 is the rule to read twice.** The epic has a behavioural exit gate and, until the gate -ratifies FIX-1430 into the role, nothing in the set runs it. An epic that wraps on ER-1, ER-2 and -ER-3 shipping would be claiming routing works rather than having seen it. +**ER-20 is the rule to read twice.** Until the gate ratifies FIX-1430 into the role, nothing in the +set runs the epic's own exit gate. Wrapping on ER-1, ER-2 and ER-3 shipping would be claiming +routing works rather than having seen it. diff --git a/spec/_epics/work-routing-package-cohesion/DECISIONS.md b/spec/_epics/work-routing-package-cohesion/DECISIONS.md index 132a0bd0c1..c5cfbe9ba2 100644 --- a/spec/_epics/work-routing-package-cohesion/DECISIONS.md +++ b/spec/_epics/work-routing-package-cohesion/DECISIONS.md @@ -2,11 +2,10 @@ [Spec](SPEC.md) · **Decisions** · [Rules](BUSINESS-RULES.md) · [Plan](PLAN.md) -The calls that sit above any single issue: what was chosen, what lost, and what each locks in for -the seven issues under it. Three are the sign-off surface. All five are **Architect locks carried -into this document**, not calls taken here — FIX-1407's *FSD Architect — EM guidance* section and -each child's own fences are the source, and its **Still open** list stays open -([below](#open)). +The calls above any single issue: what was chosen, what lost, what each locks in. All five are +**Architect locks carried into this document**, not calls taken here — FIX-1407's *FSD Architect — +EM guidance* section and each child's own fences are the source, and its **Still open** list stays +open except where the Architect has since closed an item himself ([D5](#d5)). ## The tree @@ -16,10 +15,10 @@ flowchart TD E -->|"kills W4 as a W3 floor child"| D2["D2 · soft-after for ship only"] E -->|"kills one noun for two registries"| D3["D3 · workers do, channels hold"] E -->|"kills the ambient transcript dump"| D4["D4 · the wire and the payload"] - E -->|"kills a second index of truth"| D5["D5 · inventory is a query"] + E -->|"kills a second index, and the fork between the layers"| D5["D5 · inventory is two layers"] ``` -Each edge names what the decision killed. The alternative that lost, and why, is in the card. +Each edge names what the decision killed. The alternative that lost is in the card. ## D1 · The W4 first cut is one hop: channel/org board → seat runs, and assign team seats @@ -27,21 +26,20 @@ Each edge names what the decision killed. The alternative that lost, and why, is | | | |---|---| | **Instead of** | Making the full nested cascade — channel/org board → personal board → request board → seat — the exit gate | -| **Because** | The routing claim only needs one hop to be true or false. A cascade proves the same thing three times and cannot be finished on a schedule, which is how an epic with a behavioural gate never closes | -| **Locks in** | Personal / request nested decomposition is **phase-2 inside W4**, promoted by a POC that proves need, not by default. Three children (FIX-817, FIX-1415, and FIX-1430's nested half) sit off the gate and gate nothing. Conversation, DM and transcript stay off the board entirely — they belong to the inventory and ChannelFlow | +| **Because** | The routing claim needs one hop to be true or false. A cascade proves the same thing three times and cannot be finished on a schedule, which is how an epic with a behavioural gate never closes | +| **Locks in** | Nested personal / request decomposition is **phase-2 inside W4**, promoted by a POC that proves need. Three children (FIX-817, FIX-1415, FIX-1430's nested half) sit off the gate. Conversation, DM and transcript stay off the board — they belong to the inventory and ChannelFlow | -**What would change my mind on the objective:** evidence that one hop is already possible today by -composition, so the set is documenting rather than building. Then W4 is a docs pass with an -inventory attached, and the shape is wrong. +Off the gate is not free: a phase-2 child that is still parented holds the epic's **wrap** open +after the gate is met. That is [Open](#open), not a qualification on this card. ## D2 · W4 *ship* tickets are soft-after W3; W4 is not a child of W3 | | | |---|---| -| **Instead of** | Nesting W4 under W3 as a floor child · or letting a W4 ship PR merge onto a floor still being edited | -| **Because** | Every W4 child reads the file surface W3 is still closing. Nesting makes one epic that cannot wrap; ignoring the order means a package format lands against a `tools:` fence and a `TEAM.md` layer that are still specs | -| **Locks in** | Filing, specs and **parallel POCs run now** on every child. No W4 *ship* PR merges while W3 has open children — today FIX-1377, FIX-1416, FIX-1435 and FIX-1441 ([ER-14](BUSINESS-RULES.md)). The relation is soft-after / soft-related in Linear, never parent | +| **Instead of** | Nesting W4 under W3 · or letting a W4 ship PR merge onto a floor still being edited | +| **Because** | Every W4 child reads the file surface W3 is still closing. Nesting makes one epic that cannot wrap; ignoring the order lands a package format against a `tools:` fence and a `TEAM.md` layer that are still specs | +| **Locks in** | Filing, specs and **parallel POCs run now**. The fence lifts when **every W3 child that carries an implementation is merged to main**; children completed by decision, duplicated or cancelled do not hold it ([ER-14](BUSINESS-RULES.md)). Soft-after / soft-related in Linear, never parent | ## D3 · Workers and seats *do*; channels *hold*. A board assignee key is not a Workforce seat @@ -49,112 +47,104 @@ inventory attached, and the shape is wrong. | | | |---|---| | **Instead of** | One noun for both — merging the board's assignee registry with the Workforce roster · or an assignable channel that claims work as an executor | -| **Because** | They are different lifetimes with different identity rules. A board assignee is a key on a ledger; a seat is a roster slot that mints a flow instance. Collapsing them means every board row implies a hired seat, and every seat implies a board | -| **Locks in** | The two map **by composition**, per child. No assignable-channel routing, no channel-as-claim-worker, no second `WorkerRegistry`. `defaultWorker` is triage composition and `assignTask` is composition, not new primitives. Also locked: **Agent is an opinionated default kind**, not a Layer-1 substrate type, and there are no thin/fat seat labels | +| **Because** | Different lifetimes, different identity rules. A board assignee is a key on a ledger; a seat is a roster slot that mints a flow instance. Collapsing them means every board row implies a hired seat, and every seat implies a board | +| **Locks in** | The two map **by composition**, per child. No assignable-channel routing, no channel-as-claim-worker, no second `WorkerRegistry`. `defaultWorker` and `assignTask` are composition, not new primitives. **Agent is an opinionated default kind**, not a Layer-1 substrate type, and there are no thin/fat seat labels | -## D4 · Dispatch is the wire; handoff content is the payload policy — they are not competing forks +## D4 · Dispatch is the wire; handoff content is the payload policy — not competing forks | | | |---|---| | **Instead of** | Choosing between "pass the parent session" and "pass a brief" · or a silent full parent transcript as the default | -| **Because** | One is mechanism and one is policy, and reading them as a fork is what produces both an ambient dump *and* a second runtime. A sub-agent is same-session background work with a different prompt; a worker assign is a roster seat in a **linked** session. Both need the link; only the second needs a payload rule | -| **Locks in** | Dispatch always passes `parentSessionId`, and the child binds its parent **for life at mint**. Parent history reaches a child through **opt-in tools**, never an ambient dump, and a parent's private seat history never auto-bleeds. The default payload is a required brief — `goal` / `constraints` / `acceptance` / `links` — plus channel transcript **only if both are on that channel**, and opt-in named packs. A channel session and a parent worker session may coexist with separate tool packs. **Shipped on FIX-1408**, as a decision rather than as code | +| **Because** | One is mechanism, one is policy, and reading them as a fork is what produces both an ambient dump *and* a second runtime. A sub-agent is same-session background work with a different prompt; a worker assign is a roster seat in a **linked** session. Both need the link; only the second needs a payload rule | +| **Locks in** | Dispatch always passes `parentSessionId`; the child binds its parent **for life at mint**. Parent history reaches a child by **opt-in tools**, never an ambient dump, and a parent's private seat history never auto-bleeds. The default payload is the brief — `goal` / `constraints` / `acceptance` / `links` — plus channel transcript **only if both are on that channel**. **Shipped on FIX-1408**, as a decision rather than code | + +**Linked is not nested.** A linked session is a sibling that names its parent. Nothing here +re-legitimizes nested, child or shadow sessions as a work hierarchy: +[FIX-1440](https://linear.app/fixpoint-labs/issue/FIX-1440) removed that substrate and FIX-1408 +invent-kills its return. No child reads `parentSessionId` as a session tree. -## D5 · Runtime inventory is an L2 *query* over the one shared scan, never a second registry +## D5 · Runtime inventory is **two layers**, and choosing between them is not a fork | | | |---|---| -| **Instead of** | A parallel mega-index of seats and channels · a second `WorkerRegistry` · rebuilding Graft for membership and DM lookup | -| **Because** | The shared Workforce map already exists and one scan already fills it. A second source of truth is a staleness bug with a schedule, and it is the thing three of this epic's children would each be tempted to build separately | -| **Locks in** | Inventory reads that map plus hire, `channelInstances` and open sessions. FIX-817's manifests read the same surface rather than building their own; FIX-1415's invite / find consumes it. `team.*` wildcards are a **later caller**, not part of the first ship, and do not block the channels convention. **The owner's stamped lean** is that ChannelFlow maintains an org-scoped resource — a lean, not a close: helper-vs-resource, eager-vs-lazy expansion, and DM as shared-kind-session vs special lane all stay on the [Still open](#open) list | +| **Instead of** | A second `WorkerRegistry`, a mega-loader that re-walks the tree differently, or a Graft rebuild for membership and DM lookup — **and** treating read-time compose versus an org-scoped resource as a fork to pick one of | +| **Because** | They answer different questions. The declared tree is what the files say; the live set is what is actually open. A compose helper goes stale the moment a channel mints; a live resource cannot tell you about a seat nobody has opened. Picking one produces the other by accident, badly | +| **Locks in** | **Layer 1 — declared roster, composed at read time.** The duplicated private `LabRoster` in `goals/devforce-lab/lab/host.mts` and `goals/pentest-lab/lab/host.mts` becomes one package-level export over the three existing readers (`readWorkforce`, `readResourcesDirectory`, `readChannelsDirectory`) — derived per read, no state, registers nothing. Both lab copies deleted. FIX-1405, or a thin sibling, owns it. **Layer 2 — runtime inventory** (Jake's lock, 2026-09-16): ChannelFlow updates an org-scoped resource as seats and channels open. It stays FIX-1405's **primary** runtime shape for DM find-or-create, membership and fan-out, and layer 1 does **not** replace it. `team.*` wildcards call layer 2 later, not at first ship | + +**Compose is not a registry.** Layer 1 is a function over what the readers already return; the +invent-kill on a second index of truth ([ER-12](BUSINESS-RULES.md)) is unchanged by it. + ## Who owns what -![Who owns what: a matrix of six cross-cutting rules against the seven issues in the set. The board is the work plane is built by FIX-1385 and consumed by FIX-1430. One package format is built by FIX-1394 and consumed by FIX-1430. One runtime lookup is built by FIX-1405 and consumed by FIX-1385, FIX-817 and FIX-1415. The wire and the payload is decided by FIX-1408 and consumed by FIX-1385, FIX-1394 and FIX-1430. Assignee is not a seat is decided by FIX-1385 and consumed by FIX-1394 and FIX-1430. The last row, the propagation pass inherited from the project as PR-5, has no owner in the set and is drawn as a single gap spanning every column.](figures/ownership.svg) +![Who owns what: a matrix of six cross-cutting rules against the seven issues in the set. The board is the work plane is built by FIX-1385 and consumed by FIX-1430. One package format is built by FIX-1394 and consumed by FIX-1430. Inventory in two layers is built by FIX-1405 and consumed by FIX-1385, FIX-817 and FIX-1415. The wire and the payload is decided by FIX-1408 and consumed by FIX-1385, FIX-1394 and FIX-1430. Assignee is not a seat is decided by FIX-1385 and consumed by FIX-1394 and FIX-1430. The propagation pass inherited from the project as PR-5 is built by FIX-1385. Every rule has exactly one owner.](figures/ownership.svg) -Every rule in the set has one owner, with one exception drawn as a gap: **ER-19, the project's -PR-5 propagation pass, is assigned to this epic at project altitude and to no child in it**. Read -a row to see where a decision is made, where it is built, and where it is only consumed; a -*consumes* cell is a place a child must not re-decide. FIX-1430 consumes four rows and owns none, -which is what would make it a proof. +Every rule has exactly one owner. Read a row to see where a decision is made, where it is built, +and where it is only consumed; a *consumes* cell is a place a child must not re-decide. FIX-1430 +consumes four rows and owns none, which is what makes it a proof rather than a surface. + ## Decided in review, recorded so no child reopens them -Nothing yet — this is the epic-spec's first dispatch and the PR has had no review. Cross-cutting -answers land here as they arrive, under [ER-15](BUSINESS-RULES.md). +**Round 1 · 2026-09-18**, five reviewers on +[#1905](https://github.com/fixpoint-labs/flow-state-dev/pull/1905). + +- **The inventory is two layers, and the fork is dismissed** (Architect) → [D5](#d5), + [ER-3](BUSINESS-RULES.md). Verified against the code first: `ReadWorkforceResult` is + `{workers, errors, skillErrors}` only, the three readers are separate exports, and both labs + hand-roll the same private `LabRoster` — so "one shared scan map" was intent, not an exported + thing. The compose helper's public export stays on the owner's objective gate. +- **The W3 ship fence has one release condition** (Greptile P1). Three documents stated it three + ways — *merges*, *no open children*, *close* — which diverge the moment a child closes by + decision. Now stated once everywhere: every W3 child **that carries an implementation** is merged + to main. +- **The proof consumes the package *contract*, not the implementation** (Greptile P1, sharpened by + Codex). The FIX-1394 → FIX-1430 edge is non-blocking with the condition named + ([the spec](SPEC.md#what-the-proof-consumes)). ER-2 still binds the proof: no third package shape. +- **FIX-1408's unclosed walls are owned, not orphaned.** Reuse-vs-create session policy, auto-scale + / busy-copy, hire-or-dispatch-with-parent naming, which opt-in history packs are v1, and how a + sub-agent's background work surfaces on a board row: **FIX-1394's POC owns the evidence, this + epic owns the decision**. FIX-1408 is not reopened, and no child decides one locally + ([ER-15](BUSINESS-RULES.md)). +- **FIX-1385 owns ER-19**, the project's PR-5 propagation pass — it mints the largest new Layer 2 + vocabulary surface in the set. ## What the end-state POC showed -**No end-state POC has been built.** The division into issues came from the Architect's child -ownership carve rather than from a sketch, and four of the seven children are themselves -POC-first — FIX-1394 explicitly so. If the gate wants the division tested before any of it starts, -an end-state POC on this branch is the instrument, and the question it would answer is whether the -package format and the board's assign surface are really two issues. +**None built.** The division into issues came from the Architect's ownership carve rather than a +sketch, and four of the seven children are themselves POC-first. If the gate wants the division +tested first, the question one would answer is whether the package format and the board's assign +surface are really two issues. ## Open -**1. Does the set ratify at seven — and is FIX-1430 the epic's proof?** -*(Decides: the owner. Blocks: nothing today; it decides what the epic promises and when it can -wrap.)* +**Seven children, or five?** *(Decides: the owner. Blocks: nothing today — it decides when the epic +can wrap.)* -- **The fork.** Keep all seven with two marked phase-2 and FIX-1430 adopted as the proof — or cut - the epic back to the four the body names, and let FIX-817 and FIX-1415 stand on their own? +- **The fork.** Keep all seven parented — or cut to five (FIX-1394, FIX-1405, FIX-1385, FIX-1408 + done, FIX-1430 as the proof) and re-home FIX-817 and FIX-1415 as related-not-child, dependency + links preserved? - **In plain terms.** The epic promises one behaviour: work filed for a team reaches a seat that - runs it. Three of the seven issues build that. One (FIX-1430) is the only thing that would - *demonstrate* it, and right now the epic has a behavioural promise and nothing that checks it. - The other two — a catalog an agent can read to plan, and letting a seat create and invite to - channels — are real work that serves the same area but not the promise; both tickets fence - themselves off a ship until other things land. -- **The trade-off.** Keeping seven means one epic owns the whole routing surface, its two - phase-2 children stay next to the inventory they read, and a reader sees the real backlog. It - also means a set with three of seven on the path to its own gate — the shape where a tail grows - while the proof never starts, which is exactly what W3's necessity check was watching. Cutting - to four keeps the epic honest about what it promises, and costs two issues re-homed with no - epic holding their shared dependency on FIX-1405. -- **My recommendation: keep seven, mark FIX-817 and FIX-1415 phase-2 explicitly, and adopt - FIX-1430 as the proof.** A behavioural exit gate with no child that runs it is the defect worth - fixing here; FIX-1430 already describes exactly the gate's one hop, and its own nested-board half - is already fenced to phase-2 by its Architect section. The two phase-2 children cost nothing - while they gate nothing, and splitting them from FIX-1405 would put one manifest surface in two - epics — which is what [D5](#d5) exists to prevent. -- **What would change my mind.** If you want W4 to wrap fast on the three named issues, cut it to - four: the proof can be a fifth issue filed against the gate rather than an existing showcase - adopted into the role. And if FIX-1430 is intended as a DevForce Lab artefact rather than a - framework proof, adopting it makes the epic's finish depend on a Lab's schedule. -- **If wrong.** Two issues sit in an epic that wraps before they start, and get re-parented — a - tracker edit. Or the epic wraps on three shipped surfaces with nothing having run them, which is - the more expensive half, because "work routes to a seat" would be a claim rather than a result. - -**2. Who owns FIX-1408's unclosed walls?** *(Decides: the owner, with the Architect. Blocks: -nothing today; it decides where the reuse-vs-create policy lands.)* FIX-1408 closed as **Done** -with its *locked distinctions* recorded ([D4](#d4)) and its own open walls untouched: -reuse-vs-create session policy after POC, auto-scale / busy-copy behaviour, the -hire-or-dispatch-with-parent API naming, which opt-in history packs are v1, and how a sub-agent's -background work surfaces on a board row. Those are cross-cutting — FIX-1394's and FIX-1430's POCs -both press on them — and they now have no owning issue. **Recommendation:** let FIX-1394's POC -matrix answer reuse-vs-create and record the answer here, rather than re-opening a closed ticket. -**If wrong:** the first child to hit it decides locally, which is the second-authority failure -[ER-15](BUSINESS-RULES.md) exists to prevent. - -**3. The project's PR-5 propagation pass is assigned to this epic and to no child in it.** -*(Decides: the owner. Blocks: nothing.)* The project spec -([#1818](https://github.com/fixpoint-labs/flow-state-dev/pull/1818)) names FIX-1407 as PR-5's -owner — *a new surface uses the settled Layer 2 name, never the superseded one* — checked at "the -propagation pass". No child in the set carries that work, and the epic body does not mention it. -Recorded as [ER-19](BUSINESS-RULES.md) with no owner rather than quietly dropped. It is a gap, not -a refusal; nothing in the set is blocked by it. - -## How it got here - -- **Stood up (Sep 18)** — four documents from FIX-1407's D-4 body and its Architect EM guidance, - plus each child's own fences. No product direction was added: the exit gate, the sequencing, the - vocabulary and every invent-kill are carried, and the Architect's **Still open** list is carried - open. -- **The set was written as seven, not four (Sep 18)** — FIX-817, FIX-1415 and FIX-1430 are - parented in Linear but absent from the epic body. Recorded in the set table with the split - visible, and put to the gate as Open 1 rather than resolved on the branch. -- **FIX-1408 recorded as done-by-decision (Sep 18)** — no implementation PR; its locked - distinctions became D4 and ER-4, and its unclosed walls became Open 2. + runs it. Three issues build it, one demonstrates it, one is already done. The other two — a + catalog an agent reads to plan, and letting a seat create and invite to channels — are real work + in the same area that the promise does not need. +- **The trade-off.** Not "phase-2 costs nothing", which is what an earlier draft of this document + implied and what one reviewer argued from. **The exit gate and the wrap are different moments** + ([the spec](SPEC.md#the-set--as-of-2026-09-18) has the mechanism). Keeping seven means W4 proves + what it set out to prove and then stays open, indefinitely, on two children it does not need. + Cutting to five means W4 wraps when it hits its gate. +- **My recommendation: five.** An epic that cannot close after meeting its own exit gate has the + wrong boundary, and both children fence themselves off a ship by their own Architect sections + anyway. +- **What would change my mind.** The objection to splitting is that FIX-1405 would then have two + dependent epics. That is answerable: FIX-817 and FIX-1415 are phase-2 *after* FIX-1405 lands, so + the second dependent consumes shipped code, not a concurrent epic. What would actually change my + mind is wanting one epic to hold the whole channel/board/inventory arc as a tracking unit, and + being willing to pay for it with a W4 that stays open past its gate. +- **If wrong.** Cutting when you should have kept costs two re-parented issues and a dependency + link to redraw — a tracker edit. Keeping when you should have cut costs the epic's finish: it + sits open for as long as two phase-2 children take, and every wrap check reports it as live work + when it is not. diff --git a/spec/_epics/work-routing-package-cohesion/PLAN.md b/spec/_epics/work-routing-package-cohesion/PLAN.md index 4be348e5d2..0557d8db2a 100644 --- a/spec/_epics/work-routing-package-cohesion/PLAN.md +++ b/spec/_epics/work-routing-package-cohesion/PLAN.md @@ -2,100 +2,100 @@ [Spec](SPEC.md) · [Decisions](DECISIONS.md) · [Rules](BUSINESS-RULES.md) · **Plan** -An epic plan sequences the work and says what each piece entails. It does not say how to build any -piece; that's each issue's own plan. IDs cross-reference [DECISIONS.md](DECISIONS.md) (D-n) and +An epic plan sequences the work and says what each piece entails — not how to build any piece, +which is each issue's own plan. IDs cross-reference [DECISIONS.md](DECISIONS.md) (D-n) and [BUSINESS-RULES.md](BUSINESS-RULES.md) (ER-n). ## The path -![The path: a ship fence band across the top marking W3 open with four children, then two W3 input lanes for TEAM.md and the tools fence whose specs are approved and unbuilt, then seven issue lanes against time. FIX-1408 has a done bar behind the now line at September 18. The three first-cut lanes — package cohesion, runtime inventory and channel boards — are empty ahead of the now line and marked spec and POC may start now, ship merge held by the fence. The two phase-2 lanes for catalog manifests and channel-admin sit further right, off the gate. The manager-queue lab lane is last and marked proof, unadopted. The critical path is drawn in the gutter through the inventory, the boards and the lab.](figures/path.svg) +![The path: a ship fence band across the top saying no W4 ship PR merges until every W3 child carrying an implementation is merged to main, with four W3 children still to merge, then two W3 input lanes for TEAM.md and the tools fence whose specs are approved and unbuilt, then seven issue lanes against time. FIX-1408 has a done bar behind the now line at September 18. The three first-cut lanes — package cohesion, runtime inventory and channel boards — are empty ahead of the now line and marked spec and POC may start now, ship merge held by the fence. The two phase-2 lanes for catalog manifests and channel-admin sit further right, off the gate but holding the wrap. The manager-queue lab lane is last and marked proof, unadopted. The critical path is drawn in the gutter through the inventory, the boards and the lab.](figures/path.svg) -Nine lanes: two inputs from W3 and the seven children. **One bar exists** — FIX-1408, done by -decision — and everything else is an empty lane, because nothing else has started. That is the -honest picture at the gate, and it is why the band across the top matters more than any lane: the -fence is on **merge**, not on work ([ER-14](BUSINESS-RULES.md)), so all six remaining children can -open a spec and a POC the day the objective is approved. The critical path is short and runs -inventory → boards → the proof; package cohesion is the one lane that can run beside it the whole -way. The dependency graph itself is in -[the spec](SPEC.md#how-the-issues-flow-into-each-other); this document adds time to it. +Nine lanes: two W3 inputs and the seven children. **One bar exists** — FIX-1408, done by decision. +Everything else is an empty lane, because nothing else has started. That is the honest picture at +the gate, and it is why the band across the top matters more than any lane: the fence is on +**merge**, not on work ([ER-14](BUSINESS-RULES.md)), so all six remaining children can open a spec +and a POC the day the objective is approved. The critical path runs inventory → boards → the proof. +Package cohesion runs beside it the whole way, because what the proof consumes from it is a +ratified contract rather than a shipped package ([the spec](SPEC.md#what-the-proof-consumes)). The +dependency graph is in [the spec](SPEC.md#how-the-issues-flow-into-each-other); this adds time. ## What each issue entails | Issue | Route | Consumes | Delivers | Releases | Size | |---|---|---|---|---|---| | **FIX-1408** dispatch policy | design → **closed as decided** | D4's two leans | Sub-agent vs seat-assign, `parentSessionId` at mint, opt-in history, the brief | Every other child | — | -| **FIX-1405** runtime inventory | spec → POC → impl | The shared scan · hire · `channelInstances` · open sessions (D5) | One L2 lookup for membership expand, author check, fan-out, DM find-or-create | FIX-1385 · FIX-817 · FIX-1415 · `team.*` wildcards, later | Medium | -| **FIX-1385** channel boards | spec → POC → impl | The ChannelFlow floor · the inventory (ER-3) · ER-5 | A channel holding `0..N` TaskCollections; two doors, one mutation surface | The proof · FIX-1430 | Large | -| **FIX-1394** package cohesion | **POC matrix → ratify** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both unlanded** (ER-18) · D4 | One package format, two attachment modes; the scale-back before both surfaces grow twice | Ship tickets, cut after ratify | Large | -| **FIX-817** catalog manifests · *phase-2* | spec → impl | The inventory as a reader, not a second index (D5) | One discovery shape across seats, channels, resources, tools, skills | Planning before assign | Medium | -| **FIX-1415** channel-admin · *phase-2* | spec → ratify → ship tickets | The inventory · the `tools:` fence · Collab's system-vs-dynamic lanes | Channel create / delete / invite as a **capability**, behind seat `tools:` | DevForce's "EM opens a channel" pressure | Medium | -| **FIX-1430** manager-queue lab · *proof, unadopted* | spec → goal check | ER-1 · ER-2 · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, with queue columns as views | The epic's wrap | Medium | +| **FIX-1405** runtime inventory | spec → POC → impl | The three existing readers · hire · `channelInstances` · open sessions (D5) | Two layers: a declared roster composed at read time (replacing both labs' private `LabRoster`), and the live org resource | FIX-1385 · FIX-817 · FIX-1415 · wildcards, later | Medium | +| **FIX-1385** channel boards | spec → POC → impl | The ChannelFlow floor · the inventory (ER-3) · ER-5 | A channel holding `0..N` TaskCollections; two doors, one surface. Plus the PR-5 propagation pass (ER-19) | The proof · FIX-1430 | Large | +| **FIX-1394** package cohesion | **POC matrix → ratify** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both unlanded** (ER-18) · D4 | One package format, two attachment modes. Its POC also carries the evidence for FIX-1408's open walls (ER-15) | The proof's contract at ratify; ship tickets after | Large | +| **FIX-817** catalog manifests · *phase-2* | spec → impl | FIX-1405's **approved reader contract** (ER-23), as a reader and not a second index | One discovery shape over seats, channels, resources, tools, skills | Planning before assign | Medium | +| **FIX-1415** channel-admin · *phase-2* | spec → ratify → ship tickets | The inventory · the `tools:` fence after FIX-1394 (ER-23) · Collab's system-vs-dynamic lanes | Channel create / delete / invite as a **capability**, behind seat `tools:` | DevForce's "EM opens a channel" pressure | Medium | +| **FIX-1430** manager-queue lab · *proof, unadopted* | spec → goal check | ER-1 · ER-2 as a contract · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, queue columns as views | The epic's wrap | Medium | **Size is a read, not an estimate.** Four of the seven are exploration tickets whose own scope is -the thing being explored; FIX-1394 is the clearest case — its deliverable is a POC matrix, and the -ship tickets it cuts are not in this set yet. +the thing being explored — FIX-1394 most of all, whose deliverable is a POC matrix and whose ship +tickets are not in this set yet. ## Where it is -Status lives in one place: [the set table in the spec](SPEC.md#the-set--as-of-2026-09-18). The -lanes above carry the same state as a picture of time and are redrawn when it moves. +Status lives in one place: [the set table](SPEC.md#the-set--as-of-2026-09-18). The lanes above are +the same state as a picture of time, redrawn when it moves. -**The inputs from other epics, and their verified state.** FIX-1377 (optional `TEAM.md`) and -FIX-1416 (custom tools authoring) are W3 children whose specs are **approved** — +**The inputs from other epics.** FIX-1377 (optional `TEAM.md`) and FIX-1416 (custom tools +authoring) are W3 children whose specs are **approved** — [#1819](https://github.com/fixpoint-labs/flow-state-dev/pull/1819) and -[#1890](https://github.com/fixpoint-labs/flow-state-dev/pull/1890), both carrying the -`spec approved` label, both closed unmerged — and whose code has **not landed**. Linear still reads -*In Spec Review* for both, which is [ER-16](BUSINESS-RULES.md) already failing on the two -dependencies this epic cares most about. Treat the approved specs as the contract (ER-18); treat -the Linear state as stale. +[#1890](https://github.com/fixpoint-labs/flow-state-dev/pull/1890), both carrying `spec approved`, +both closed unmerged — and whose code has **not landed**. Both are In Development as of 2026-09-18. +Write against the approved specs and name them (ER-18). They are dashed edges in the graph because +a contract is what you get, not because the code is there. ## What unblocks what, from here 1. **The objective is approved** → all six live children may open a spec and a POC immediately. - Nothing in the set waits on anything else to *start*; the fence is on merge (ER-14). -2. **FIX-1405's inventory lands** → FIX-1385 can assign a **team** seat rather than a literal id, - and FIX-817 and FIX-1415 stop being blocked on a lookup that doesn't exist. This is the one - edge that genuinely sequences. -3. **FIX-1385's board surface lands** → the proof can run. Until then ER-20 has nothing behind it. -4. **W3's last child merges** → the ship fence lifts and W4 impl PRs may merge. Four remain: - FIX-1377, FIX-1416, FIX-1435, FIX-1441. -5. **FIX-1394's POC matrix is ratified** → and only then are its ship tickets cut (ER-8). Those - tickets are new children of this epic and will need the set table and this path redrawn. -6. **If the gate adopts FIX-1430 as the proof** → ER-20 gets an owner and the epic gains a - finishing condition. If it doesn't, the epic needs one filed, or it wraps on a claim. + Nothing waits on anything else to *start*; the fence is on merge (ER-14). +2. **FIX-1405's reader contract is approved** → FIX-817 may start specifying against it (ER-23). + Before that, two children would design one discovery API in parallel. +3. **FIX-1405's inventory lands** → FIX-1385 can assign a **team** seat rather than a literal id. + The one edge that genuinely sequences. +4. **FIX-1385's board surface lands** → the proof can run. Until then ER-20 has nothing behind it. +5. **W3's last implementing child merges to main** → the ship fence lifts. Children completed by + decision, duplicated or cancelled do not hold it (ER-14). +6. **FIX-1394's POC matrix is ratified** → ER-2 becomes a contract the proof can hold itself to, + and only then are ship tickets cut (ER-8). Those are new children and will need the set table + and this path redrawn. +7. **If the gate adopts FIX-1430 as the proof** → ER-20 gets an owner and the epic gains a + finishing condition. If not, one needs filing, or the epic wraps on a claim. ## Coordination seams to watch | Seam | Between | Rule | |---|---|---| -| The inventory surface | FIX-1405 and FIX-817 | 1405 owns it; 817 reads it. Whether the inventory resource **is** the manifest or only feeds a tool is an open wall — neither child answers it alone (D5, ER-13) | -| The seat's `tools:` fence | FIX-1394, FIX-1415, and FIX-1416 outside the epic | Three edits pressing on one fence. Neither W4 child widens past seat `tools:`, and both write against FIX-1416's approved spec rather than around it (ER-18) | -| The seat's prompt compose | FIX-1394 and FIX-1377 outside the epic | The package format composes onto framework default → team → seat. It does not invent a fourth layer, and it does not merge the team's text into the seat's | -| The board's claim / assign surface | FIX-1385 and FIX-1430 | 1385 owns it; the lab consumes it. Cross-flow seats must declare the **same logical board** (`boardId` + ledger + own same-flow gating dispatcher) — whether that is a file-convention teach is FIX-1385's open wall, and the lab does not answer it | -| `goals/devforce-lab/` | FIX-1430 and FIX-1426 (Done, outside the set) | 1426 is the coding seam only and does not grow into the manager-queue showcase | -| The Workforce docs section | FIX-1385, FIX-1405, FIX-1430 | All three teach boards. Whichever lands second links rather than repeats, and none teaches ahead of its reader | +| The inventory surface | FIX-1405 and FIX-817 | 1405 owns both layers; 817 reads them, and specs only after 1405's reader contract is approved (ER-23) | +| The seat's `tools:` fence | FIX-1394, FIX-1415, and FIX-1416 outside the epic | Three edits on one fence, settled in that order (ER-23). Neither W4 child widens past seat `tools:` | +| The seat's prompt compose | FIX-1394 and FIX-1377 outside the epic | The package format composes onto framework default → team → seat. No fourth layer; the team's text does not merge into the seat's | +| The board's claim / assign surface | FIX-1385 and FIX-1430 | 1385 owns it, the lab consumes it. Cross-flow seats declare the **same logical board** (`boardId` + ledger + own same-flow gating dispatcher) — whether that is a file-convention teach is FIX-1385's open wall, not the lab's | +| The Workforce docs section | FIX-1385, FIX-1405, FIX-1430 | All three teach boards. Whichever lands second links rather than repeats | ## Not children, deliberately -[FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351) (**W3**) — soft-after for ship, -soft-related otherwise, never re-parented (D2). -[FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341) (**Collab RC**) — parked; FIX-1415 -shapes a capability, not a reopen. -[FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359) (**OOTB agent kind**) — shipped, -parallel history, not reopened inside W4. -Also consumed, not owned: [FIX-1388](https://linear.app/fixpoint-labs/issue/FIX-1388) (Door B), -[FIX-1393](https://linear.app/fixpoint-labs/issue/FIX-1393) (the `tools:` fence), -[FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426) (DevForce first slice, Done), -[FIX-1440](https://linear.app/fixpoint-labs/issue/FIX-1440) (nested-session removal — cited, not a -ship child), [FIX-1410](https://linear.app/fixpoint-labs/issue/FIX-1410) (D-12 DevForce Lab), -[FIX-1373](https://linear.app/fixpoint-labs/issue/FIX-1373) (board-worker vocab park) and -[FIX-957](https://linear.app/fixpoint-labs/issue/FIX-957) (durable task collections). +Soft-after for ship, never re-parented: [FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351) +(**W3**, D2). Parked: [FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341) (**Collab RC**) — +FIX-1415 shapes a capability, not a reopen. Shipped and not reopened inside W4: +[FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359) (OOTB agent kind), +[FIX-1440](https://linear.app/fixpoint-labs/issue/FIX-1440) (nested-session removal — cited as a +fence, never a ship child). Consumed, not owned: +[FIX-1388](https://linear.app/fixpoint-labs/issue/FIX-1388), +[FIX-1393](https://linear.app/fixpoint-labs/issue/FIX-1393), +[FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426), +[FIX-1410](https://linear.app/fixpoint-labs/issue/FIX-1410), +[FIX-1373](https://linear.app/fixpoint-labs/issue/FIX-1373), +[FIX-957](https://linear.app/fixpoint-labs/issue/FIX-957). ## Wrap -When ER-20 holds: run the lessons pass over the set's review rounds; dispatch the docs polish over -the Workforce pages the children each edited in isolation — it owes the sentence nobody else will -write, that a board **assignee** and a Workforce **seat** are two different things that map by -composition (ER-5, ER-22). Then settle whether PR-5's propagation pass ran or is handed on -([ER-19](BUSINESS-RULES.md)), refresh the set table and the path one last time, and close the epic -PR unmerged. +When ER-20 holds **and every remaining child is Linear-terminal** — not the same moment while +FIX-817 and FIX-1415 are parented, which is the set question ([Open](DECISIONS.md#open)): run the +lessons pass over the set's review rounds, then dispatch the docs polish over the Workforce pages +the children each edited in isolation. It owes the sentence nobody else will write — that a board +**assignee** and a Workforce **seat** are two different things that map by composition (ER-5, +ER-22). Confirm FIX-1385 ran the PR-5 propagation pass (ER-19), refresh the set table and the path +one last time, and close the epic PR unmerged. diff --git a/spec/_epics/work-routing-package-cohesion/SPEC.md b/spec/_epics/work-routing-package-cohesion/SPEC.md index 086acd005e..c78df2d70f 100644 --- a/spec/_epics/work-routing-package-cohesion/SPEC.md +++ b/spec/_epics/work-routing-package-cohesion/SPEC.md @@ -9,65 +9,56 @@ architecture cohesion | A team that… | Today | After this epic's first cut | |---|---|---| -| **has a piece of work for a team of seats** | Nowhere shared to put it. Work is dispatched by hand, from code, one seat at a time | It goes on the channel's board. A seat claims it or is assigned it, and runs it | -| **writes one package for a seat** | Two overlapping systems — the seat's own prompt file and skills — and no answer to which one a tool belongs in | One package format, two attachment modes: always-on for a seat, opt-in for a library | -| **asks what seats and channels exist right now** | Opaque strings. `engineering.*` cannot expand, and a seat asking another seat has no way to find their DM | One runtime lookup answers membership, author checks, fan-out and DM find-or-create | -| **hands work from one seat to another** | Invents an ambient transcript dump, or a second runtime | A named brief, a linked session bound to its parent for life, and parent history only through opt-in tools | - -**Why now.** W3 made a Workforce *describable* — workers, channels, resources and skills all -declare in one tree. Nothing yet says how **work reaches** one of those declared seats. Four -surfaces are missing or disagree: the package story (a seat's prompt file versus skills), what -exists at runtime (no inventory, so membership is an opaque string), what *assign* means (a -sub-agent, a roster seat, a board row — all three, unreconciled), and where work lives at all. -That gap is why a declared team is still a demo. This epic closes the routing half of Goal 1: W3 -proved a seat is *described*, W4 proves a seat is *given work*. It does not move the corpus-wide -goal count on its own — the lead measure moves when a child ships a goal check. +| **has work for a team of seats** | Nowhere shared to put it. Dispatched by hand, from code, one seat at a time | It goes on the channel's board. A seat claims it or is assigned it, and runs it | +| **writes one package for a seat** | Two overlapping systems — the seat's prompt file and skills — and no answer to which one a tool belongs in | One package format, two attachment modes: always-on for a seat, opt-in for a library | +| **asks what seats and channels exist** | Opaque strings. `engineering.*` cannot expand, and a seat cannot find another seat's DM | A declared roster read from the tree, plus a live org resource for what is actually open | +| **hands work from one seat to another** | Invents an ambient transcript dump, or a second runtime | A brief, a linked session bound to its parent for life, and parent history only by opt-in tools | + +**Why now.** W3 made a Workforce *describable*. Nothing yet says how **work reaches** one of those +declared seats: no shared place to file it, no runtime answer to "which seats exist", and three +unreconciled meanings of *assign*. That gap is why a declared team is still a demo. W3 proved a +seat is described; W4 proves a seat is given work. It does not move the corpus-wide goal count on +its own — the lead measure moves when a child ships a goal check. ## What's in the box -![What's in the box: the W4 first cut is a channel or org board that routes work to a seat run and assigns team seats — the exit gate, one hop — plus one package format with two attachment modes, a runtime inventory of seats and channels as one L2 query, and the shipped dispatch and session policy. Composed in by the app: custom kinds as flow factories, and defaultWorker triage. A phase-2 panel inside W4, off the exit gate, holds the nested personal and request board cascade, channel-admin verbs, catalog manifests and the manager-queue lab's nested half. Below a fence, what is not built: no Agent, Channel, Team, MessageBoard, TeamFlow or SessionBoard L1 type, no Collab RC, no silent parent transcript, no merged board-assignee and seat registry, no team wildcards as first ship, no second WorkerRegistry or mega-loader, no assignable-channel routing, no BoardFlow as the only mint, no Graft rebuild.](figures/end-state.svg) +![What's in the box: the W4 first cut is a channel or org board that routes work to a seat run and assigns team seats — the exit gate, one hop — plus one package format with two attachment modes, a runtime inventory in two layers (a declared roster composed at read time and a live org resource ChannelFlow updates), and the shipped dispatch and session policy. Composed in by the app: custom kinds as flow factories, and defaultWorker triage. A phase-2 panel inside W4, off the exit gate but holding the wrap, holds the nested personal and request board cascade, channel-admin verbs, catalog manifests and the manager-queue lab's nested half. Below a fence, what is not built: no Agent, Channel, Team, MessageBoard, TeamFlow or SessionBoard L1 type, no Collab RC, no silent parent transcript, no nested or shadow session substrate as a work hierarchy, no merged board-assignee and seat registry, no team wildcards as first ship, no second WorkerRegistry or mega-loader, no assignable-channel routing, no BoardFlow as the only mint, no Graft rebuild.](figures/end-state.svg) -Everything inside the box is what a team gets once the first cut lands. The **exit gate is the top -row and only the top row**: a channel or org board routes work to a seat run, and can assign team -seats ([D1](DECISIONS.md#d1)). The strip under it is phase-2 *inside* W4 — named so nobody reads -it as refused, and fenced off the gate so the epic can finish. The bottom strip is what the set -refuses to build, and every item in it was named out by the Architect rather than forgotten. +The **exit gate is the top row and only the top row** ([D1](DECISIONS.md#d1)). The strip under it +is phase-2 *inside* W4 — named so nobody reads it as refused, and fenced off the gate so the gate +is reachable on a schedule. Off the gate is not free; see below. The bottom strip is what the set +refuses to build, every item named out by the Architect rather than forgotten. ## The set · as of 2026-09-18 -This table is the live one. It is refreshed on the epic PR as issues move; the plan and the -figures point here rather than repeating it. - -**The epic body names four.** Three more were parented after it was written, and whether they -belong is the one thing the objective gate has to settle — see [sign-off](#sign-off) and -[Open 1](DECISIONS.md#open). +The live table. Refreshed on the epic PR as issues move; the plan and the figures point here. | Issue | What it delivers | Why the set needs it | Status | |---|---|---|---| -| FIX-1408 | Dispatch / session policy — sub-agent is same-session background, worker assign is a roster seat in a **linked** session, `parentSessionId` always at mint, history by opt-in tools | The wire everything else routes over. Locked as decision, not code | **Done** · closed 2026-09-18 · no impl PR ([below](#a-note-on-fix-1408)) | -| FIX-1394 | Package cohesion — one package format, two attachment modes (seat always-on, library opt-in). POC-first | Half the epic's title. Two overlapping surfaces grow board and tool sugar twice until this settles | Not started · Backlog | -| FIX-1405 | Runtime inventory of seats + channels; lean is an org-scoped resource ChannelFlow maintains | Assigning a **team** seat needs to know which seats exist. Without it the board can route to a literal id and nothing else | Not started · Backlog | -| FIX-1385 | Channel workstreams + boards — a channel holds `0..N` TaskCollections; channel actions and `taskTools` are two doors on one surface | **The exit gate's surface.** It is the board in "channel/org board → seat runs" | Not started · Todo | -| FIX-817 | Collection / catalog manifests + agent introspection — one discovery shape over seats, channels, resources, tools, skills | Planning *before* assign. Not on the path to the exit gate; it reads the same inventory FIX-1405 builds | Not started · Todo · `Open Question` | -| FIX-1415 | Channel-admin capability — create / delete / invite behind the seat's `tools:` fence | Design only until Collab park and inventory land, by its own Architect fence. Off the exit gate | Not started · Todo | -| FIX-1430 | Manager-queue lab — a coordinator seat owns a channel + workstream board, assigns to linked seats, and shows queue state as **views** over existing task status | The only child shaped like a **proof** of the exit gate. Nothing else in the set runs it end to end | Not started · Todo | - -1 done · 0 in flight · 6 not started. Four are the epic body's set; three (FIX-817, FIX-1415, -FIX-1430) were parented later and the body does not name them. - -**Is seven really four?** Seven is **three plus a proof plus two phase-2** — and the counting is -the gate's call, not mine. FIX-1394, FIX-1405 and FIX-1385 are the first cut the body names. -FIX-1430 is the only child that would *run* the exit gate, and a set with a behavioural gate and -nothing that runs it is missing its proof, so my recommendation is to adopt it as one. FIX-817 and -FIX-1415 both sit behind FIX-1405 and both fence themselves off a ship; kept as phase-2, they gate -nothing and keep one shared surface with its owner. The full trade, including the arm where two of -them leave the epic, is [Open 1](DECISIONS.md#open). +| FIX-1408 | Dispatch / session policy: sub-agent is same-session background, worker assign is a roster seat in a **linked** session, `parentSessionId` at mint, history by opt-in tools | The wire everything else routes over | **Done** · 2026-09-18 · no impl PR ([below](#a-note-on-fix-1408)) | +| FIX-1394 | One package format, two attachment modes. POC-first | Half the epic's title. Two overlapping surfaces grow board and tool sugar twice until this settles | Not started · Backlog | +| FIX-1405 | Inventory in two layers: a declared roster composed from the existing readers, and the live org resource ChannelFlow updates | Assigning a **team** seat needs to know which seats exist. Without it a board routes to a literal id and nothing else | Not started · Backlog | +| FIX-1385 | A channel holding `0..N` TaskCollections; channel actions and `taskTools` as two doors on one surface. Also the PR-5 propagation pass ([ER-19](BUSINESS-RULES.md)) | **The exit gate's surface** — the board in "channel/org board → seat runs" | Not started · Todo | +| FIX-817 | Catalog manifests + agent introspection: one discovery shape over seats, channels, resources, tools, skills | Planning *before* assign. Off the exit gate, but **holds the wrap** | Not started · Todo · `Open Question` | +| FIX-1415 | Channel create / delete / invite as a capability, behind the seat's `tools:` fence | Off the exit gate by its own fence, but **holds the wrap** | Not started · Todo | +| FIX-1430 | Manager-queue lab: a coordinator seat owns a channel board, assigns to linked seats, shows queue state as **views** over existing task status | The only child shaped like a **proof** of the exit gate | Not started · Todo | + +1 done · 0 in flight · 6 not started. + +**Four, or seven?** The epic body names four — FIX-1408, FIX-1394, FIX-1405, FIX-1385. FIX-817, +FIX-1415 and FIX-1430 were parented afterwards and the body does not mention them. That is what the +objective gate has to settle, and it is not bookkeeping, because **the epic's exit gate and its +wrap are different moments.** The wrap term in `.agents/workflows/epic-wake.js` requires every child +to be Linear-terminal, merged or `DONE`, and `TERMINAL_LINEAR` matches only done / closed / +cancelled / duplicate / dropped / won't-do. A Backlog or Todo child is none of those. So a phase-2 +child does not gate the exit gate — it holds the epic open past it, indefinitely. Both arms priced: +[Open](DECISIONS.md#open). -**A note on FIX-1408.** It went Backlog → Done on 2026-09-18 with no implementation PR: what -shipped is the **decision**, recorded as [D4](DECISIONS.md#d4) and [ER-4](BUSINESS-RULES.md). Its -own open walls are not closed by that and now have no owning issue -([Open 2](DECISIONS.md#open)). +**A note on FIX-1408.** Backlog → Done on 2026-09-18 with no implementation PR: what shipped is the +**decision** ([D4](DECISIONS.md#d4), [ER-4](BUSINESS-RULES.md)). Its own open walls are not closed +by that — FIX-1394's POC owns the evidence and this epic owns the decision +([decided in review](DECISIONS.md#decided-in-review)). ## How the issues flow into each other @@ -79,50 +70,57 @@ flowchart LR DP --> PK["FIX-1394 · package cohesion"] INV["FIX-1405 · runtime inventory"] -->|"which seats exist to assign to"| BD["FIX-1385 · channel boards"] BD -->|"the exit gate's surface"| LAB["FIX-1430 · manager-queue lab"] - PK -->|"one package a seat carries"| LAB - INV -->|"the same manifest shape"| CAT["FIX-817 · catalog manifests · phase-2"] + PK -.->|"the ratified contract, not the ship"| LAB + INV -->|"reader contract closes first"| CAT["FIX-817 · catalog manifests · phase-2"] INV -.->|"invite and find"| ADM["FIX-1415 · channel-admin · phase-2"] T416 -.-> ADM classDef done stroke-width:2px class DP done ``` -A solid edge is what one issue hands the next and blocks until it lands. **Dashed edges from the -left are inputs from W3 that have not landed** — FIX-1377 and FIX-1416 are spec-approved and still -to be built, so no W4 child may write as though they exist ([ER-18](BUSINESS-RULES.md)). The -dashed edge into FIX-1415 is soft: its own fence holds it for Collab and inventory, so it waits -rather than blocks. A heavy border is done. Every node is filed, so the set carries no -placeholders. +A **solid** edge blocks: what it carries does not exist until it lands. A **dashed** edge does not, +and its label says what makes starting safe anyway — an approved spec to write against (the two W3 +inputs, [ER-18](BUSINESS-RULES.md)), a ratified contract rather than a shipped surface (below), or +a fence the child holds itself to (FIX-1415). A heavy border is done. Every node is filed. + + +**The proof consumes package cohesion's contract, not its implementation.** FIX-1430's seats can +carry instructions and tools on today's surfaces — both existing labs already do. What the proof +may not do is invent a *third* shape the ratified format would contradict; that is +[ER-2](BUSINESS-RULES.md) binding it. Binding the proof to a shipped package instead would hang the +exit gate on ship tickets that are cut only after a POC ratify and do not exist yet — the +unreachable gate [D1](DECISIONS.md#d1) exists to prevent. ## What stays as it is -- **Collab RC ([FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341)).** Dynamic rooms, - membership mutate and the roster stay parked. FIX-1415 shapes a capability surface; it does not - reopen Collab. -- **The W3 floor ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351)).** A sibling epic, +- **Collab RC ([FIX-1341](https://linear.app/fixpoint-labs/issue/FIX-1341))** — parked. FIX-1415 + shapes a capability surface; it does not reopen Collab. +- **The W3 floor ([FIX-1351](https://linear.app/fixpoint-labs/issue/FIX-1351))** — a sibling epic, soft-*after* for ship and never a parent ([D2](DECISIONS.md#d2)). -- **The OOTB agent kind ([FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359)).** Shipped, - parallel history. Package cohesion flags drift against it; it is not redesigned from zero. -- **The board's claim system.** One task board, one claim system. `assignee` stays optional and - `TaskStatus` does not grow ([ER-11](BUSINESS-RULES.md)). -- **`goals/devforce-lab/` ([FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426), Done).** - The coding seam only. It does not grow into FIX-1430's showcase. +- **Session substrate ([FIX-1440](https://linear.app/fixpoint-labs/issue/FIX-1440))** — linked + dispatch does not re-legitimize nested, child or shadow sessions as a work hierarchy + ([D4](DECISIONS.md#d4)). +- **The OOTB agent kind ([FIX-1359](https://linear.app/fixpoint-labs/issue/FIX-1359))** — shipped. + Package cohesion flags drift against it; it is not redesigned from zero. +- **The board's claim system** — `assignee` stays optional, `TaskStatus` does not grow + ([ER-11](BUSINESS-RULES.md)). +- **`goals/devforce-lab/` ([FIX-1426](https://linear.app/fixpoint-labs/issue/FIX-1426))** — the + coding seam only. It does not grow into FIX-1430's showcase. ## Sign off -1. **[D1](DECISIONS.md#d1) · The exit gate is channel/org board → seat runs / assign team seats. - The nested personal/request cascade is phase-2.** If wrong: the epic either finishes on a cut - too thin to prove routing, or never finishes because the gate kept moving. -2. **[Open 1](DECISIONS.md#open) · Does the set ratify at seven, and is FIX-1430 the proof?** - Three issues were parented after the body was written. Two of them (FIX-817, FIX-1415) are - phase-2 by their own fences; the third (FIX-1430) is the only child shaped to run the gate, and - the epic currently has a behavioural exit gate with **nothing that proves it**. The full - six-part ask is in [DECISIONS.md → Open](DECISIONS.md#open). -3. **[D2](DECISIONS.md#d2) · W4 *ship* PRs are soft-after W3; filing, specs and POCs run now.** If - wrong: either W4 ships onto a floor that moves under it, or four children idle for a floor that - was never going to block them. - -**Open: two.** Whether the set is seven, and who owns FIX-1408's unclosed walls -([DECISIONS.md → Open](DECISIONS.md#open)). The rules every child obeys are in -[BUSINESS-RULES.md](BUSINESS-RULES.md); the order the work runs in, and what each issue entails, -is [PLAN.md](PLAN.md). +1. **[D1](DECISIONS.md#d1) · The exit gate is channel/org board → seat runs / assign team seats; + the nested cascade is phase-2.** If wrong: the epic finishes on a cut too thin to prove routing, + or never finishes because the gate kept moving. +2. **[Open](DECISIONS.md#open) · Seven children, or five?** Keeping FIX-817 and FIX-1415 parented + means W4 reaches its exit gate and then stays open until they are done. Re-homing them as + related-not-child means W4 wraps at its gate. **Recommendation: five.** If wrong: either the + epic sits open for months after proving what it set out to prove, or two issues lose the epic + holding their shared dependency on FIX-1405. +3. **[D2](DECISIONS.md#d2) · W4 *ship* PRs are soft-after W3; filing, specs and POCs run now.** The + fence lifts when **every W3 child that carries an implementation is merged to main**; children + completed by decision, duplicated or cancelled do not hold it. If wrong: either W4 ships onto a + floor that moves under it, or six children idle for a floor that was never going to block them. + +**Open: one** — the set. Rules: [BUSINESS-RULES.md](BUSINESS-RULES.md). Sequencing and what each +issue entails: [PLAN.md](PLAN.md). diff --git a/spec/_epics/work-routing-package-cohesion/figures/end-state.svg b/spec/_epics/work-routing-package-cohesion/figures/end-state.svg index 12a1a55646..e2b87fb5b7 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/end-state.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/end-state.svg @@ -1,5 +1,5 @@ - + diff --git a/spec/_epics/work-routing-package-cohesion/figures/ownership.svg b/spec/_epics/work-routing-package-cohesion/figures/ownership.svg index 7cc773156b..4f99f153be 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/ownership.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/ownership.svg @@ -1,5 +1,5 @@ + aria-label="Who owns what: a matrix of six cross-cutting rules against the seven issues in the set. The board is the work plane is built by FIX-1385 and consumed by FIX-1430. One package format is built by FIX-1394 and consumed by FIX-1430. Inventory in two layers is built by FIX-1405 and consumed by FIX-1385, FIX-817 and FIX-1415. The wire and the payload is decided by FIX-1408 and consumed by FIX-1385, FIX-1394 and FIX-1430. Assignee is not a seat is decided by FIX-1385 and consumed by FIX-1394 and FIX-1430. The propagation pass inherited from the project as PR-5 is built by FIX-1385. Every rule has exactly one owner."> - WHO OWNS WHAT — EVERY RULE HAS ONE OWNER, EXCEPT THE LAST + WHO OWNS WHAT — EVERY RULE HAS EXACTLY ONE OWNER FIX-1385 @@ -57,7 +55,7 @@ - ER-3 · one runtime lookup + ER-3 · inventory in two layers consumes BUILDS @@ -91,8 +89,13 @@ ER-19 · the propagation pass - - no owner in the set — inherited from the project as PR-5, assigned to the epic itself + BUILDS + + + + + + @@ -100,8 +103,6 @@ the one owner — decides or builds it consumes it, and must not re-decide it - - no owner — a gap, not a refusal - FIX-1430 consumes four rows and owns none, which is what would make it the proof + FIX-1430 consumes four rows and owns none, which is what makes it the proof diff --git a/spec/_epics/work-routing-package-cohesion/figures/path.svg b/spec/_epics/work-routing-package-cohesion/figures/path.svg index 324759eb3c..4f3af3ab1f 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/path.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/path.svg @@ -1,5 +1,5 @@ + aria-label="The path: a ship-fence band under the axis saying the W3 floor is open with four children and no W4 ship PR merges until every W3 child carrying an implementation is merged to main. Below it, nine lanes against time with a now line at September 18. Two input lanes from W3, TEAM.md and the tools fence, carry bars that stop at the now line and read spec approved and unbuilt. FIX-1408 carries the only done bar, done by decision. The three first-cut lanes — runtime inventory, channel boards and package cohesion — are empty lanes ahead of the now line and read spec and POC may start now. The two phase-2 lanes, catalog manifests and channel-admin, are empty and read off the gate but holding the wrap. The last lane, the manager-queue lab, is empty and reads the proof, unadopted. The critical path is drawn in the gutter through the inventory, the boards and the lab."> From 8c620bc22bd1dc09b9bb173c13bad4b34317d5ff Mon Sep 17 00:00:00 2001 From: Claude Date: Sat, 19 Sep 2026 15:15:07 +0000 Subject: [PATCH 14/21] spec(epic): #1928 and #1929 merged; ER-20 still NOT RUN (FIX-1407) Four of five children are now complete. The gate is not. - FIX-1405 terminal: #1920, #1923 and #1928 all merged (#1928 merged by the owner at 15:08:04Z, merge commit 8428614). - FIX-1430 terminal: #1929 merged (15:11:47Z, merge commit 45ef6d7) off head 78bc4f8. - FIX-1394 is the only child still open; its matrix #1921 never merges and its ratify waits on the documents question. ER-20 is unchanged and still reads NOT RUN. Merging the proof did not run the proof. The spec now says plainly that every child of the set can be complete while the thing the set exists to prove has never been observed, so nobody reads "everything merged" as "the epic is done". The wrap stays enforced against the rule as written, and the two things in front of it are both the owner's: the documents question, and a machine with an inference key. ER-14, recorded precisely and deliberately narrowly: - The owner merged #1928 through a live fence - FIX-1435 has no branch and no PR at all, verified here, not remembered. That act SETTLES #1928. It is not a breach and nothing about it needs remediation. - It is NOT an answer about the rule. He has said nothing. The re-gate keeps stating the narrowing recommendation as the epic's and unanswered, and the merge is not cited as approval anywhere. - What is recorded as decided is only the epic's own OPERATING DEFAULT (DECISIONS.md#fence-default): treat the fence as narrowed to per-PR file overlap, read from his action rather than his words, countermandable by him at any time. - ER-14's row text is byte-identical again. Only the note around it moved. - ER-14's remaining bite is named: FIX-1394's eventual ship PR, and the wrap. Path figure redrawn and looked at in both themes: FIX-1405 and FIX-1430 become merged bars, and FIX-1430's runs straight into the red ER-20 NOT RUN gate - which is the picture worth having right now. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01PxDE76gD3BxXnPPjT7BT1T --- .../BUSINESS-RULES.md | 29 ++++++---- .../DECISIONS.md | 58 ++++++++++++------- .../work-routing-package-cohesion/PLAN.md | 54 +++++++++-------- .../work-routing-package-cohesion/SPEC.md | 39 ++++++++----- .../figures/path.svg | 13 ++--- 5 files changed, 117 insertions(+), 76 deletions(-) diff --git a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md index 393f618f57..f898f1af63 100644 --- a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md +++ b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md @@ -43,16 +43,25 @@ FIX-1351's 22 children, exactly two are non-terminal: **FIX-1435** (`Todo`) and carries an implementation** — it collapses two copies of the slot-descent walk in the resources loaders, framework code W4 children read. FIX-1449 is Atlas HTML staleness and carries none. -**So the fence has not lifted. Three W4 ship PRs merged under it** — FIX-1385's -[#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922) and FIX-1405's -[#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920) and -[#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923), all shipping -`packages/workforce` with changesets — **and a fourth, -[#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928), is open and parked by it.** -Recorded, not re-decided: ER-14 derives from [D2](DECISIONS.md#d2), which the owner ratified at the -objective gate, so the epic cannot narrow it here. Whether the fence's unit should be the package or -the files a PR touches is **[a pending re-gate](DECISIONS.md#er-14-re-gate)** with the owner, where -the overlap evidence for #1928 is written out. +**So the fence has not lifted, and four W4 ship PRs have merged under it** — FIX-1385's +[#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922), FIX-1405's +[#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920), +[#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923) and +[#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928), all shipping +`packages/workforce` with changesets. **The owner merged the last of them himself**, on 2026-09-19. +That settles #1928; it is not a change to this rule, and he has said nothing about one. + +**What is still fenced.** Only two things in this set remain subject to ER-14: **FIX-1394's +eventual ship PR** — not started, and behind the documents question and the ratify (ER-8) — +and **the [wrap term](PLAN.md#wrap)**, which cannot complete while a W4 ship PR is parked. Nothing +else in the set is still waiting on this fence. + +**How the epic behaves meanwhile.** ER-14 derives from [D2](DECISIONS.md#d2), which the owner +ratified, so the epic cannot narrow the rule — and it has not. What it has adopted is an +[operating default](DECISIONS.md#fence-default): treat the fence as narrowed to per-PR file +overlap, read from the owner's action rather than his words, and **countermandable by him at any +time**. Whether the rule itself narrows is still **[a pending re-gate](DECISIONS.md#er-14-re-gate)** +with the owner, where the recommendation is the epic's and unanswered. ## How the set is run diff --git a/spec/_epics/work-routing-package-cohesion/DECISIONS.md b/spec/_epics/work-routing-package-cohesion/DECISIONS.md index 081101ce19..cc43b7f381 100644 --- a/spec/_epics/work-routing-package-cohesion/DECISIONS.md +++ b/spec/_epics/work-routing-package-cohesion/DECISIONS.md @@ -128,13 +128,17 @@ rather than a matrix row, and **FIX-1385 provides the surface ER-20 stands on wi | **A package is authored in Markdown**, as the new file, scoped to **instructions and tools** — and the format **may not be designed disk-only** | The owner, 2026-09-19, on FIX-1394's first ratify fork. His own direction is the load-bearing half: packages an LLM writes are **stored as a resource, never saved to disk**, so a format readable only off the filesystem is wrong on arrival. *Don't collapse* is no longer a live result | [ER-2](BUSINESS-RULES.md) · [Open](#open) | | ER-3 means **(a)** a contrast with the declared tree — **not (b)** relocating ChannelFlow's post fence onto an org resource | A channel's `members` is already live, read on the refusal path in the session the post lands in; (b) would put one fact in two places and tax every post | ER-3 · ER-12 | | **Drain width is 1.** Work queues behind a busy seat; a busy seat never takes a second row concurrently | Closes FIX-1408's *auto-scale / busy-copy* wall on FIX-1430's evidence, the way [ER-15](BUSINESS-RULES.md) says such a wall closes: the lab ran at **both** widths and wrote the comparison out, rather than a number being picked. **Reversible** — the `MANAGER_QUEUE_DRAIN_WIDTH` knob stays, which is how the other case remains reachable | FIX-1430's `lab/README.md` · [Open](#open), one wall lighter | +| **The epic treats the W3 ship fence as narrowed to per-PR file overlap**, from 2026-09-19 | **An operating default, not a ruling, and not the owner's.** He merged #1928 through the live fence; the epic reads that action — not any statement, because there was none — as licence to stop parking W4 ship PRs whose diffs do not touch an unlanded W3 child's files. Recorded so the epic's behaviour is predictable and attributable rather than improvised per PR. **[The re-gate](#er-14-re-gate) is still open and still the owner's**, and **he can countermand this in a sentence**, at which point the default reverts to the rule as written | [ER-14](BUSINESS-RULES.md)'s note · [the re-gate](#er-14-re-gate) | | **Two of FIX-1430's approved rules had their stated mechanism refuted by the code.** Both texts are corrected; both rules stand and are graded | **BR-5:** a row does run in its own session, with a parent bound at mint and none of the coordinator's transcript — but the parent is the **draining seat's** session, not the filer's, because the dispatch that mints the child happens at the seat, after the filing. **BR-11:** `adoptLapsedLease` (`packages/orchestration/src/task-board/task-entry.ts`) **renews** a lapsed lease so a successor *can* take the row back; `StaleTaskClaimError` fires only when a reclaim genuinely won. The substance — a lapsed row is really back in the queue — now grades the stronger true claim: the next drain takes it and runs it on the seat whose own file answers for the desk it was filed for | ER-4 · FIX-1430's `BUSINESS-RULES.md` and `lab/README.md` | The first four rows are round 1 (2026-09-18, five reviewers on [#1905](https://github.com/fixpoint-labs/flow-state-dev/pull/1905)); the rest are 2026-09-19, raised -up from a child rather than decided locally ([ER-15](BUSINESS-RULES.md) working). **The last two -rows are a child's PR correcting the epic** — the same mechanism as the cross-spec pass above it, -pointed at a fact rather than a scope. None reopens an approved decision. +up from a child rather than decided locally ([ER-15](BUSINESS-RULES.md) working). **The drain-width +and refuted-mechanism rows are a child's PR correcting the epic** — the same mechanism as the +cross-spec pass above them, pointed at a fact rather than a scope. **[The fence +default](#fence-default) is the one row that is neither**: it is how the epic behaves while a +question with the owner stays open, and it decides nothing that question will decide. None of them +reopens an approved decision. **Two stay falsifiable.** ER-3's reading flips on evidence that a channel's `members` is *not* current on the refusal path; the ER-15 split flips if those walls turn out to carry package content @@ -178,26 +182,35 @@ the wrap open indefinitely. **What would change my mind:** if skill-context hone ### Does the W3 ship fence narrow to per-PR overlap? — a pending re-gate *(Decides: the owner, and only the owner — this document records him ratifying **the ship fence as -written** at the objective gate, so narrowing it is a re-gate, not an epic call. Blocks: whether -[#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) may merge. The recommendation -below is **mine**, and unanswered.)* +written** at the objective gate, so narrowing it is a re-gate, not an epic call. Blocks: now +**FIX-1394's eventual ship PR** and the [wrap term](PLAN.md#wrap); no longer #1928. The +recommendation below is **the epic's**, and **unanswered**.)* **The facts.** #1928 publishes `@flow-state-dev/workforce` and carries a changeset, so it is a **ship PR** by [ER-14](BUSINESS-RULES.md)'s letter. FIX-1435 — a W3 child carrying an -implementation, hoisting the duplicated slot-descent walk in the resources loaders — is still -`Todo`, so the fence has not lifted and it parks a finished PR. **By substance the overlap is -empty:** #1928's 13 paths sit under `packages/workforce/src/channel/`, `src/inventory/`, `test/`, -the README, `apps/docs/` and a changeset — **none** under `packages/workforce/src/loader/`, where -FIX-1435 lands. Same package, disjoint files. - -**My recommendation: narrow ER-14 to per-PR overlap** — a W4 ship PR is fenced only when its diff -touches files an unlanded W3 child will edit. The rule's own stated risk is W4 shipping onto -unlanded W3 *code*, and a fence that parks on package identity rather than file overlap costs merges -it was never aimed at. **What would change my mind:** evidence that FIX-1435's hoist changes the -*behaviour* of a loader #1928's code calls — then the package is the right unit and #1928 waits. -**What being wrong costs:** a W4 change merged onto a floor still moving under it, so a conflict or -a silent behavioural change surfaces on `main` rather than in a rebase. **Nothing is narrowed -here:** ER-14's row stands exactly as written, and three W4 ship PRs have already merged under it. +implementation, hoisting the duplicated slot-descent walk in the resources loaders — is `Todo` with +**no branch and no PR at all**, so the fence was live. **By substance the overlap is empty:** +#1928's 13 paths sit under `packages/workforce/src/channel/`, `src/inventory/`, `test/`, the README, +`apps/docs/` and a changeset — **none** under `packages/workforce/src/loader/`, where FIX-1435 +lands. Same package, disjoint files. + +**The owner merged #1928 through the fence** on 2026-09-19 (`merged_by: jhoffner`, merge commit +`8428614`). That is his act and it **settles #1928** — nothing about that PR is outstanding and +nothing here asks for it to be undone. **It is not an answer to this question.** He has said +nothing about the rule, and a merge is not a rewrite: reading one as the other is how a fence gets +narrowed by accident and nobody can say when. + +**The recommendation stands, and stays the epic's: narrow ER-14 to per-PR overlap** — a W4 ship PR +is fenced only when its diff touches files an unlanded W3 child will edit. ER-14's own stated risk +is W4 shipping onto unlanded W3 *code*, and a fence that parks on package identity rather than file +overlap costs merges it was never aimed at. **What would change my mind:** evidence that FIX-1435's +hoist changes the *behaviour* of a loader a W4 child's code calls — then the package is the right +unit. **What being wrong costs:** a W4 change merged onto a floor still moving under it, so a +conflict or a silent behavioural change surfaces on `main` rather than in a rebase. **Nothing is +narrowed here:** ER-14's row stands exactly as written. + +**What the epic does in the meantime** is [its own operating default](#fence-default), not a +decision on this question — and the owner can countermand it in a sentence. **Why the three above are here and the three below are not.** Those are business calls the owner owns and nothing else can settle; the three below are engineering walls that *building* answers @@ -215,8 +228,9 @@ auto-scale / busy-copy — closed on 2026-09-19 as [drain width is 1](#drain-wid | **A sub-agent's background work on a board row** | Whether same-session background work is visible on the row at all, and as what | **Unowned, and orphaned.** It was FIX-1385's as the owner of the row; FIX-1385 closed without building a coordinator assign, so nothing left in the set will force it. [ER-11](BUSINESS-RULES.md) binds whoever takes it — a view, never a new status. **Named out at [the wrap](PLAN.md#wrap)**, so the epic cannot close by confirming a pass nobody ran | **What being wrong costs.** The bet was that FIX-1385 would force three of these; it did not. The -two still settleable ride on FIX-1430 — so if the proof runs and nobody reads what it reports, they -leave W4 unanswered with no child left to force them. The third has no owner at all. +two still settleable ride on FIX-1430 — **which has now merged**, so the evidence is on `main` to be +read rather than waited for, and **this is the last cheap moment**: no child is left to force them, +and once the epic wraps nobody is looking. The third has no owner at all. **A second unowned residue, from FIX-1385.** A seat naming a board id **no channel minted** is not diff --git a/spec/_epics/work-routing-package-cohesion/PLAN.md b/spec/_epics/work-routing-package-cohesion/PLAN.md index df03a7f8f7..9d7d2a66a2 100644 --- a/spec/_epics/work-routing-package-cohesion/PLAN.md +++ b/spec/_epics/work-routing-package-cohesion/PLAN.md @@ -12,25 +12,26 @@ which is each issue's own plan. IDs cross-reference [DECISIONS.md](DECISIONS.md) Nine lanes: two W3 inputs, the **five** children D6 ratified, and — under the divider — the two [related issues](SPEC.md#related-not-children) that still consume FIX-1405. Every lane carries a -bar, and every child of the set now has a PR. The lane the tracker holds and this figure does not is +bar, and four of the five are merged. The lane the tracker holds and this figure does not is [FIX-1451](SPEC.md#the-sixth-child) — deliberately, until it is confirmed a child or re-homed. -**The critical path ran boards → the proof, and both halves are behind us.** FIX-1385 merged on -2026-09-19 and FIX-1430's build followed it onto -[#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929). What is in front of the epic now -is not a build: it is [ER-20's own run](SPEC.md#er-20-has-not-run), which needs a model credential -this epic has never had. The inventory and package cohesion are *beside* the critical path, not -behind it. +**The critical path ran boards → the proof, and all of it is behind us.** FIX-1385 merged on +2026-09-19 and FIX-1430's build followed it onto `main` the same day +([#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929)). **What is in front of the epic +is not work**: it is [ER-20's own run](SPEC.md#er-20-has-not-run), which needs a model credential +this epic has never had. That is why the proof's lane ends in a gate box rather than at its bar — +the code is done and the gate is open. Package cohesion is *beside* the critical path, not behind +it. ## What each issue entails | Issue | Route | Consumes | Delivers | Releases | Size | |---|---|---|---|---|---| | **FIX-1408** dispatch policy | design → **closed as decided** | D4's two leans | Sub-agent vs seat-assign, `parentSessionId` at mint, opt-in history, the brief | Every other child | — | -| **FIX-1405** runtime inventory | spec → POC → impl → **in review** | The three existing readers · hire · `channelInstances` · open sessions (D5) | Two layers: a declared roster composed at read time (replacing both labs' private `LabRoster`), and the live org resource | FIX-817, on its approved reader contract (ER-23) · FIX-1415 · `team.*` wildcards, later. **Not FIX-1385** | Medium · PR-A and S4 merged, S5–S7 open on [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) | +| **FIX-1405** runtime inventory | spec → POC → **impl, merged** | The three existing readers · hire · `channelInstances` · open sessions (D5) | Two layers: a declared roster composed at read time (replacing both labs' private `LabRoster`), and the live org resource | FIX-817, on its approved reader contract (ER-23) · FIX-1415 · `team.*` wildcards, later. **Not FIX-1385** | Medium · all three PRs on `main` — PR-A [#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920), S4 [#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923), S5–S7 [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) | | **FIX-1385** channel boards | spec → POC → **impl, merged** | The ChannelFlow floor · ER-5 · ER-3, which it may not re-decide and reads only **if** the roster has landed | A channel holding `0..N` TaskCollections; two doors, one surface. Plus the PR-5 name check over its own diff (ER-19) | The proof · FIX-1430 — **released 2026-09-19** | Large · shipped in one PR, not four | | **FIX-1394** package cohesion | **POC matrix → ratify** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both landed** at `d8e4c99`, and the code wins over either spec (ER-18) · D4 | One format as a Markdown file, instructions and tools, two attachment modes, **not disk-only** (ER-2); *documents in v1* still open. Its POC also carries the evidence for **one** of FIX-1408's walls (ER-15) | Ship tickets, once the ratify completes (ER-8) — **not** the proof | Large | -| **FIX-1430** manager-queue lab · *the proof* | spec → impl → **in review**, goal check **NOT RUN** | ER-1 · **ER-2 as a fence, not an input** · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, queue columns as views. Plus the drain-width comparison ER-15 called for ([width 1](DECISIONS.md#drain-width)) | The epic's wrap, **once ER-20 runs** | Medium · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) | +| **FIX-1430** manager-queue lab · *the proof* | spec → **impl, merged** — but the goal check is **NOT RUN** | ER-1 · **ER-2 as a fence, not an input** · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, queue columns as views. Plus the drain-width comparison ER-15 called for ([width 1](DECISIONS.md#drain-width)) | The epic's wrap, **once ER-20 runs** — merging did not release it | Medium · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) merged | **Size is a read, not an estimate.** Three of the five are exploration tickets whose own scope is the thing being explored — FIX-1394 most of all, whose ship tickets are not in this set yet. @@ -55,12 +56,17 @@ the graph because they do not block, not because the code is missing. edge**: FIX-1385 reads the declared roster only *if it has landed*, and its BR-7 carries the caller's `assignee` verbatim rather than resolving a seat ([ER-5](BUSINESS-RULES.md)). 4. **FIX-1385's implementation landed** ([#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922)) - → the proof's build ran. That was the set's only true blocking edge, and it is discharged. -5. **ER-20 runs on the real path** → the epic can wrap. This is now the only thing in front of it, - and it is blocked on an environment, not on work: [the gate has not run](SPEC.md#er-20-has-not-run). + → the proof's build ran, and has since merged too. The set's only true blocking edge is + discharged, and **four of the five children are on `main`**. +5. **ER-20 runs on the real path** → the epic can wrap. Nothing is in front of it but this, and it + is blocked on an **environment**, not on work: every line of code it needs is merged, and + [the gate has not run](SPEC.md#er-20-has-not-run). 6. **W3's last implementing child merges** → the ship fence lifts (ER-14). That is **FIX-1435**, - still `Todo`. Three W4 ship PRs merged under the fence and **#1928 is parked by it** — whether - the fence narrows is [a re-gate with the owner](DECISIONS.md#er-14-re-gate), not an epic call. + still `Todo` with no branch and no PR. Four W4 ship PRs have merged under the live fence, the + last of them **#1928, by the owner himself** — which settles that PR without answering the rule. + What is still fenced is **FIX-1394's eventual ship PR** and the [wrap](#wrap). Whether the rule + narrows is [a re-gate with the owner](DECISIONS.md#er-14-re-gate), not an epic call; how the + epic behaves until he answers is [an operating default](DECISIONS.md#fence-default). 7. **FIX-1394's ratify completes** → ship tickets are cut (ER-8). Half is answered; *documents in v1* is the rest. Those tickets are new children and will need this path redrawn. It does **not** unblock the proof ([the spec](SPEC.md#what-the-proof-consumes)). @@ -72,9 +78,9 @@ the graph because they do not block, not because the code is missing. | The inventory surface | FIX-1405 and FIX-817 *(related, not a child)* | 1405 owns both layers; 817 reads them, and specs only after 1405's reader contract is approved (ER-23) | | The seat's `tools:` fence | FIX-1394 · FIX-1415, outside the set · FIX-1416, **landed** | Three edits on one fence, in that order (ER-23). None widens past seat `tools:`, and FIX-1416's merged code is the fence's actual shape, not its spec | | The seat's prompt compose | FIX-1394 and FIX-1377 | Composes framework default → team → seat. No fourth layer; the team's text does not merge into the seat's | -| The board's claim / assign surface | FIX-1385 (**merged**) and FIX-1430 (**#1929**) | 1385 owned it and shipped it; the lab consumes what landed. Two of 1385's rules about that surface were read back by the lab and [corrected](DECISIONS.md#refuted-mechanisms) rather than worked around. What 1385 did **not** settle is named in [DECISIONS](DECISIONS.md#the-cross-flow-board-residue) and has no owner | +| The board's claim / assign surface | FIX-1385 (**merged**) and FIX-1430 (**#1929, merged**) | 1385 owned it and shipped it; the lab consumes what landed. Two of 1385's rules about that surface were read back by the lab and [corrected](DECISIONS.md#refuted-mechanisms) rather than worked around. What 1385 did **not** settle is named in [DECISIONS](DECISIONS.md#the-cross-flow-board-residue) and has no owner | | The same surface, again | FIX-1385 and FIX-1415 *(related, not a child)* | 1415's invite verbs land on the channel 1385 hangs a board on. Re-homing moved the wrap, not the collision | -| One channel flow factory, two PRs | FIX-1385 (**merged**) and FIX-1405's **S5–S7 ([#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928))** | Both edit `defineChannelFlow` and the binder. **Discharged in the order it named:** FIX-1385 shipped whole first, so #1928 is built on top of it rather than racing it, and its own deviations note where the landed code won. Board *declaration* stays built-in-kind-only; channel *inventory* is every kind's | +| One channel flow factory, two PRs | FIX-1385 (**merged**) and FIX-1405's **S5–S7 ([#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928), merged)** | Both edit `defineChannelFlow` and the binder. **Closed:** FIX-1385 shipped whole first and #1928 landed on top of it rather than racing it, its own deviations noting where the landed code won. Board *declaration* stays built-in-kind-only; channel *inventory* is every kind's | | The Workforce docs section | FIX-1385, FIX-1405, FIX-1430 | All three teach boards. Whichever lands second links rather than repeats | ## Not children, deliberately @@ -99,13 +105,15 @@ When **[ER-20](BUSINESS-RULES.md) holds** and every remaining child is terminal. **Terminal is read from GitHub, not from Linear.** A child is terminal when its **implementation PR is merged or closed**, or when this epic-spec records its work as **done by decision rather than as code** — which is FIX-1408, and the set table is where that is recorded. Linear is the mirror; a -merge is what makes code shipped. **Two children are terminal: FIX-1408 by decision, FIX-1385 by -merge.** FIX-1405 and FIX-1430 have open PRs; FIX-1394's ratify is half answered. - -**ER-20 has not run, so the wrap has not opened.** Merging #1929 makes FIX-1430 terminal and still -leaves the gate unrun — [the row reads `NOT RUN`](SPEC.md#er-20-has-not-run) and the term is -enforced against the rule as written. The last step before the wrap is somebody running that check -with a model credential and the verdict log saying PASS. +merge is what makes code shipped. **Four children are terminal: FIX-1408 by decision, FIX-1385, +FIX-1405 and FIX-1430 by merge.** Only FIX-1394 is not — its ratify is half answered. + +**ER-20 has not run, so the wrap has not opened.** #1929 merged, which made FIX-1430 terminal and +left the gate exactly as unrun as before — [the row reads `NOT RUN`](SPEC.md#er-20-has-not-run) and +the term is enforced against the rule as written. **Terminal children are not a met gate**, and +this set is now the clearest possible case of the difference: everything the epic builds is on +`main`, and nobody has watched it work. The last step before the wrap is somebody running that +check with a model credential and the verdict log saying PASS. **Which children count is the other open part.** Linear lists [six](SPEC.md#the-sixth-child) against D6's ratified five, and this term reads *every remaining child*. Until FIX-1451 is confirmed in or diff --git a/spec/_epics/work-routing-package-cohesion/SPEC.md b/spec/_epics/work-routing-package-cohesion/SPEC.md index be27d0f12e..0b8b69b1b3 100644 --- a/spec/_epics/work-routing-package-cohesion/SPEC.md +++ b/spec/_epics/work-routing-package-cohesion/SPEC.md @@ -37,22 +37,29 @@ document:** read the spec branch's head, or the Linear document on the issue. |---|---|---|---| | FIX-1408 | Dispatch / session policy: sub-agent is same-session background, worker assign is a roster seat in a **linked** session, `parentSessionId` at mint, history by opt-in tools | The wire everything else routes over | **Done** · 2026-09-18 · no impl PR ([below](#a-note-on-fix-1408)) | | FIX-1394 | **One package format as a Markdown file**, scoped to instructions and tools, two attachment modes, and **not disk-only** ([ER-2](BUSINESS-RULES.md)). POC-first | Half the epic's title. Two overlapping surfaces grow board and tool sugar twice until this settles | **Ratify · half answered** · matrix [#1921](https://github.com/fixpoint-labs/flow-state-dev/pull/1921) open, never merges · authorship [decided](DECISIONS.md#decided-in-review), documents still open | -| FIX-1405 | Inventory in two layers: a declared roster composed from the existing readers, and the live org resource ChannelFlow updates | The epic's third promise in its own right — *which seats and channels exist* — and what `team.*` addressing calls later ([ER-7](BUSINESS-RULES.md)). It does **not** gate the boards | **In review** · PR-A [#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920) and S4 [#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923) **merged** · S5–S7 [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) **open**, CI green, mergeable, findings closed | +| FIX-1405 | Inventory in two layers: a declared roster composed from the existing readers, and the live org resource ChannelFlow updates | The epic's third promise in its own right — *which seats and channels exist* — and what `team.*` addressing calls later ([ER-7](BUSINESS-RULES.md)). It does **not** gate the boards | **Done** · 2026-09-19 · PR-A [#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920), S4 [#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923) and S5–S7 [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) all **merged** — terminal by merge | | FIX-1385 | A channel holding `0..N` TaskCollections; channel actions and `taskTools` as two doors on one surface. Plus the PR-5 name check over its own diff ([ER-19](BUSINESS-RULES.md)) | **The exit gate's surface** — the board in "channel/org board → seat runs" | **Done** · 2026-09-19 · [#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922) **merged** — one full-scope PR, not the spec's four ([the seam it discharged](PLAN.md#coordination-seams-to-watch)) | -| FIX-1430 | Manager-queue lab: a coordinator seat owns a channel board, assigns to linked seats, shows queue state as **views** over existing task status | The **proof** of the exit gate, and the owner of [ER-20](BUSINESS-RULES.md) | **In review** · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **open**, CI green, mergeable, findings closed · **the gate itself [has not run](#er-20-has-not-run)** | +| FIX-1430 | Manager-queue lab: a coordinator seat owns a channel board, assigns to linked seats, shows queue state as **views** over existing task status | The **proof** of the exit gate, and the owner of [ER-20](BUSINESS-RULES.md) | **Done** · 2026-09-19 · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **merged** — but **the gate it exists to run [has not run](#er-20-has-not-run)** | -**All four specs are approved and every child now has its PR.** 2 done · 3 in review · 0 not -started. That is the whole set, and it is not the epic's finish. +**Four of five children are complete; one is open and blocked.** FIX-1408 done by decision, and +FIX-1385, FIX-1405 and FIX-1430 all merged to `main` on 2026-09-19. **FIX-1394 is the only child +still open** — its matrix [#1921](https://github.com/fixpoint-labs/flow-state-dev/pull/1921) never +merges, and its ratify waits on the owner's documents question. + +**Read that as four of five, not as the epic being done.** Every child in this set can be complete +while the thing the set exists to prove has never been observed — which is exactly the state W4 is +in now. **[ER-20](BUSINESS-RULES.md) is OPEN. Its row reads `NOT RUN`.** The acceptance goal calls a real model and no environment reachable from this epic has an inference key, so the check fails loudly by name rather than degrading to a scripted filer. Everything around it is green — routing, sessions, the waiting row, refusal by name — with controls that go red at the leg they name. **The wrap term -is enforced against ER-20 as written:** W4 does not wrap until that row says PASS. Merging -[#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) would not change that; a merged -proof whose gate never ran is exactly the *claiming routing works rather than having seen it* ER-20 -exists to prevent. +is enforced against ER-20 as written:** W4 does not wrap until that row says PASS. +[#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **has since merged, and it did +not change this.** A merged proof whose gate never ran is exactly the *claiming routing works rather +than having seen it* that ER-20 exists to prevent — and it is now the literal state of the set, so +the only thing standing between W4 and its wrap is a machine with an inference key. **The tracker has moved to six.** [FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451) — a @@ -153,10 +160,14 @@ writes are **stored as a resource rather than saved to disk**, so the format may disk-only ([decided](DECISIONS.md#decided-in-review), binding through [ER-2](BUSINESS-RULES.md)). -**Open: three with you, three with the epic.** **With you:** *are documents in v1?* — parked on +**Two things are between W4 and its wrap, and both are yours.** *Are documents in v1?* — parked on [#1921](https://github.com/fixpoint-labs/flow-state-dev/pull/1921), recommendation unchanged -(*leave them out*); **whether FIX-1451 belongs under this epic** ([above](#the-sixth-child)); and a -**[pending re-gate on ER-14](DECISIONS.md#er-14-re-gate)** — #1928 is a ship PR under a fence that -has not lifted, and narrowing that fence reopens what you ratified as item 3 above, so only you can -do it. **With the epic:** FIX-1408's three remaining session-policy walls -([Open](DECISIONS.md#open)) — none blocks a start. +(*leave them out*); it is the only thing holding FIX-1394. And **a machine with an inference key**, +which is the whole of what [ER-20](#er-20-has-not-run) is waiting for. Nothing in the set is waiting +on the epic. + +**Also open with you, blocking nothing today:** **whether FIX-1451 belongs under this epic** +([above](#the-sixth-child)), and a **[pending re-gate on ER-14](DECISIONS.md#er-14-re-gate)** — you +merged #1928 through the fence, which settles that PR but says nothing about the rule; narrowing it +reopens what you ratified as item 3 above, so only you can do it. **With the epic:** FIX-1408's +three remaining session-policy walls ([Open](DECISIONS.md#open)). diff --git a/spec/_epics/work-routing-package-cohesion/figures/path.svg b/spec/_epics/work-routing-package-cohesion/figures/path.svg index 379566ab30..398fe22def 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/path.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/path.svg @@ -1,5 +1,5 @@ + aria-label="The path: a ship-fence band under the axis saying FIX-1435 has no pull request and is still Todo, so the W3 ship fence has not lifted, and that four W4 ship PRs merged under the live fence, the last of them pull request 1928 by the owner himself. Below it, nine lanes against time with a now line at September 19. Two input lanes from W3, TEAM.md and the tools fence, carry bars that reach the now line and read landed on main on September 19. Then the five lanes of the set. FIX-1408 carries a done bar behind the now line, done by decision. Every remaining lane opens with a short spec-approved bar at the now line. FIX-1405 inventory then carries one merged bar running to the end of its lane: done, with pull requests 1920, 1923 and 1928 all merged. FIX-1385 channel boards carries a merged bar running to the end of its lane: Done, pull request 1922 merged, one full-scope PR rather than the four its plan named. FIX-1394 package cohesion carries an open-PR bar reading matrix pull request 1921, authorship decided, documents still open. FIX-1430, the manager-queue lab, carries its spec bar, then a merged bar for pull request 1929, and then a dashed red gate box reading ER-20 NOT RUN — the code is merged and the exit gate's own run still needs a model credential this epic does not have, which is the one thing left before the epic can wrap. Under a divider, two related lanes, catalog manifests and channel-admin, read related, not in the set: they consume FIX-1405 and hold no wrap, and catalog manifests waits for FIX-1405's approved spec. The critical path is drawn in the gutter through the boards and the lab only; the inventory lane sits beside it, not behind it."> From 5cf936498222f439a5248bcbe4c61f46d0552f57 Mon Sep 17 00:00:00 2001 From: Claude Date: Sat, 19 Sep 2026 16:27:47 +0000 Subject: [PATCH 16/21] =?UTF-8?q?spec(epic):=20the=20set=20is=20seven=20?= =?UTF-8?q?=E2=80=94=20FIX-1451=20was=20a=20child=20all=20along=20(FIX-140?= =?UTF-8?q?7)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The epic held a standing note saying FIX-1451's parenting should be cleared. The ticket was read directly instead of recalled, and it contradicts the note: parentId FIX-1407, a ## Parent section naming this epic ("soft encounter under package cohesion, not a ship-gate on FIX-1394"), and Architect guidance reading "ship-able honesty fix under FIX-1407; do not nest under FIX-1359". Filed by the owner at 12:58. The parenting is deliberate and documented; the note was wrong and the write was never made. This is a correction, not a re-gate. Nobody changed their mind — a documented parenting was correctly described by the tracker and incorrectly described here. Recorded as such, with the defect named: twice on 2026-09-19 a remembered decision about set membership disagreed with the ticket (FIX-817 one way, FIX-1451 the other), and both were caught by reading before acting. FIX-1451 is a Bug, so it skips the spec and enters at implementation with its PR as the review surface, dispatched on fix/FIX-1451-allowed-tools-honesty. It holds the wrap like any child: the wrap's tail is now three open children (FIX-1394's ratify, FIX-1381's spec, FIX-1451's fix) plus ER-20. ER-14's re-gate note names its fix conditionally — "if it ships a package" — because the changeset call has not been made. ER-20 is unchanged: still OPEN, still NOT RUN, still the rule the wrap is enforced against. ER-14's row text is byte-identical for the fourth refresh running. Figures: path.svg gains a seventh set lane and grows to 388; rendered and checked in light and dark. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01PxDE76gD3BxXnPPjT7BT1T --- .../BUSINESS-RULES.md | 15 ++-- .../DECISIONS.md | 79 +++++++++++-------- .../work-routing-package-cohesion/PLAN.md | 79 +++++++++++-------- .../work-routing-package-cohesion/SPEC.md | 62 ++++++++++----- .../figures/path.svg | 53 +++++++------ 5 files changed, 167 insertions(+), 121 deletions(-) diff --git a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md index b91f7d5293..25c5da78a9 100644 --- a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md +++ b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md @@ -51,11 +51,12 @@ loaders, framework code W4 children read. FIX-1449 is Atlas HTML staleness and c `packages/workforce` with changesets. **The owner merged the last of them himself**, on 2026-09-19. That settles #1928; it is not a change to this rule, and he has said nothing about one. -**What is still fenced.** Three things: **FIX-1394's ship tickets**, now reachable since the ratify -unblocked (ER-8); **FIX-1381's eventual implementation**, which edits `packages/workforce` and so -will be a ship PR when it exists; and **the [wrap term](PLAN.md#wrap)**, which cannot complete while -a W4 ship PR is parked. The fence's bite therefore **grew** with the set rather than shrinking as -the merges landed. +**What is still fenced.** **FIX-1394's ship tickets**, now reachable since the ratify unblocked +(ER-8); **FIX-1381's eventual implementation**, which edits `packages/workforce` and so will be a +ship PR when it exists; **[FIX-1451](SPEC.md#the-seventh-child)'s fix, if it ships a package** — the +changeset call on that bug has not been made, so it is fenced conditionally, not certainly; and **the +[wrap term](PLAN.md#wrap)**, which cannot complete while a W4 ship PR is parked. The fence's bite +therefore **grew** with the set rather than shrinking as the merges landed. **How the epic behaves meanwhile.** ER-14 derives from [D2](DECISIONS.md#d2), which the owner ratified, so the epic cannot narrow the rule — and it has not. What it has adopted is an @@ -69,8 +70,8 @@ with the owner, where the recommendation is the epic's and unanswered. | # | Rule | Because | |---|---|---| | ER-15 | A cross-cutting question is commented **up** on the epic PR, never decided locally. FIX-1408's unclosed walls are **epic-owned**, and their current standing is in [Open](DECISIONS.md#open) | A local answer is a second authority. A wall closes on evidence a child produced and folded up here — never on a child's own say-so | -| ER-16 | A child's Linear state is mirrored the moment it changes. **Holding as of 2026-09-19**, the egress outage over and the queued backlog written | The epic wake derives blocked-by from Linear, so stale state silently mis-orders the set. The wrap term is **not** re-coupled to Linear by this — it stays [derived from GitHub](PLAN.md#terminal), because what makes code shipped is a merge. What Linear *does* decide is which children exist: the set is **six** since the owner pulled FIX-1381 in ([D6](DECISIONS.md#d6-extended)), and the tracker lists a **seventh**, [FIX-1451](SPEC.md#the-unconfirmed-child), that was never confirmed | -| ER-17 | Every child routes to *spec* by default; only a `Bug` label re-routes | Fail-closed. The original five carry `Design` + `Feature`, and **FIX-1381 is routing to spec** like any other child — its spec is being written now | +| ER-16 | A child's Linear state is mirrored the moment it changes. **Holding as of 2026-09-19**, the egress outage over and the queued backlog written | The epic wake derives blocked-by from Linear, so stale state silently mis-orders the set. The wrap term is **not** re-coupled to Linear by this — it stays [derived from GitHub](PLAN.md#terminal), because what makes code shipped is a merge. What Linear *does* decide is which children exist, and it decided it twice on 2026-09-19: the set is **seven** — FIX-1381 pulled in by the owner ([D6](DECISIONS.md#d6-extended)), and [FIX-1451](SPEC.md#the-seventh-child) confirmed as having been parented all along ([the correction](DECISIONS.md#d6-seventh)). **No unconfirmed remainder.** The second of those is the rule earning its place: the tracker was right and this spec was wrong | +| ER-17 | Every child routes to *spec* by default; only a `Bug` label re-routes | Fail-closed. The original five carry `Design` + `Feature`, and **FIX-1381 is routing to spec** like any other child — its spec is being written now. **[FIX-1451](SPEC.md#the-seventh-child) is the first child to take the other branch:** labelled `Bug`, so it skips the spec and enters at implementation with its PR as the review surface. There is no spec gate to wait for on it, and its absence is not a gap | | ER-18 | A child depending on FIX-1377 or FIX-1416 writes against the **landed code** at `d8e4c99` and names it. Where a W3 spec and that code disagree, **the code wins** | **FIX-1416 shipped stricter than its spec**, which said a seat's own `blocks/` folder is callable without being listed anywhere. `resolveDeclaredTools` (`packages/workforce/src/hire.ts`) promotes only names declared in `tools:`, so a child carrying the spec's sentence forward builds against behaviour that does not exist | | ER-19 | The project's **PR-5** — a new surface uses the settled Layer 2 name, never the superseded one — is met here as **a check over this set's own diff**, owned by **FIX-1385**. It is **not** a repo-wide rename | The repo-wide pass waits on the project's vocabulary lock (**PD-4** — running it early pays the rename twice), so it stays the project's and has **no owner inside W4**: named out here rather than dropped, so nobody wraps this epic believing it ran | | ER-23 | **A shared contract's spec is approved before the next child specs against it.** FIX-1405's reader contract — **`readDeclaredRoster(root)`** on `@flow-state-dev/workforce/loader`, collecting rather than throwing, joining the layers on **`id` alone** — closes before FIX-817 starts; the `tools:` fence settles across FIX-1416 → FIX-1394 → FIX-1415 in that order | Two issues designing one discovery API at once is how you ship two of them. It outlives FIX-817 leaving the set: the dependency is on the contract, not the parent | diff --git a/spec/_epics/work-routing-package-cohesion/DECISIONS.md b/spec/_epics/work-routing-package-cohesion/DECISIONS.md index bc8833e8ac..9ac3349ee4 100644 --- a/spec/_epics/work-routing-package-cohesion/DECISIONS.md +++ b/spec/_epics/work-routing-package-cohesion/DECISIONS.md @@ -24,7 +24,7 @@ flowchart TD E -->|"kills one noun for two registries"| D3["D3 · workers do, channels hold"] E -->|"kills the ambient transcript dump"| D4["D4 · the wire and the payload"] E -->|"kills a second index, and the fork between the layers"| D5["D5 · inventory is two layers"] - E -->|"kills an epic that outlives its own gate"| D6["D6 · the set is six, extended from five"] + E -->|"kills an epic that outlives its own gate"| D6["D6 · the set is seven, grown from five"] ``` Each edge names what the decision killed. The alternative that lost is in the card. @@ -86,13 +86,13 @@ Layer 1's **public export is approved** (objective gate), and FIX-1405 specified `readDeclaredRoster(root)` — the contract [ER-23](BUSINESS-RULES.md) blocks FIX-817 on. -## D6 · The set is six children; two related issues are re-homed rather than parented +## D6 · The set is seven children; two related issues are re-homed rather than parented | | | |---|---| -| **Instead of** | Keeping all seven parented — an epic that meets its own exit gate and then stays open, indefinitely, on two children it does not need | +| **Instead of** | Keeping FIX-817 and FIX-1415 parented as well — an epic that meets its own exit gate and then stays open, indefinitely, on two children it does not need | | **Because** | **The exit gate and the wrap are different moments.** The wrap term requires every child to be [terminal](PLAN.md#terminal), and a phase-2 child that has never been built is not. An epic that cannot close after proving what it set out to prove has the wrong boundary | -| **Locks in** | **Originally five:** FIX-1394, FIX-1405, FIX-1385, FIX-1408 (done), and **FIX-1430 adopted as the proof**, which gives [ER-20](BUSINESS-RULES.md) its owner. **Extended to six** — see below. FIX-817 and FIX-1415 are **related-not-child**, keeping their FIX-1405 dependency and this set's rules — FIX-817 still waits for FIX-1405's approved spec ([ER-23](BUSINESS-RULES.md)), and neither may re-decide [ER-3](BUSINESS-RULES.md) | +| **Locks in** | **Originally five:** FIX-1394, FIX-1405, FIX-1385, FIX-1408 (done), and **FIX-1430 adopted as the proof**, which gives [ER-20](BUSINESS-RULES.md) its owner. **Now seven** — FIX-1381 pulled in and FIX-1451 confirmed, both below. FIX-817 and FIX-1415 are **related-not-child**, keeping their FIX-1405 dependency and this set's rules — FIX-817 still waits for FIX-1405's approved spec ([ER-23](BUSINESS-RULES.md)), and neither may re-decide [ER-3](BUSINESS-RULES.md) | **Re-homed is not descoped.** Both stay filed and keep their dependency edges; what changed is which epic's wrap they hold: none. @@ -113,10 +113,25 @@ in front of him. **It is a cheaper child than a cold one:** its direction is alr thin allowlist, Ask 3 org `ro` automatic and `rw` by permission — so what is outstanding is its spec, not its shape. -**FIX-1451 is a separate, still-open question**, and this amendment does not touch it: it was -parented but never confirmed, so the tracker now holds **seven** against a set of six. A refresh -does not reopen an owner call, and does not make one either — it stays [an open -question](#open). + +**Corrected 2026-09-19 — the set is seven, and FIX-1451 was always the seventh.** The ticket was +read directly instead of from memory, and it contradicts what this document said: `parentId: +FIX-1407`, a `## Parent` section naming this epic — *"soft encounter under package cohesion, not a +ship-gate on FIX-1394"* — and Architect guidance reading *"ship-able honesty fix under FIX-1407; do +not nest under FIX-1359"*, filed by the owner at 12:58. **This is not a new call and not a re-gate.** +Nobody changed their mind; a documented parenting was correctly described by the tracker and +incorrectly described here. The standing note said to clear it, the note was wrong, and the write +was never made — which is the whole of why it cost nothing. + +**It is a `Bug`, so no spec.** It enters at implementation with its PR as the review surface +([ER-17](BUSINESS-RULES.md)), and it **holds the wrap** like any child. Counting FIX-1381, the +wrap's tail is now three open children plus ER-20. + +**The defect is worth more than the correction.** Twice on 2026-09-19 a remembered decision about +set membership disagreed with the ticket — [FIX-817](SPEC.md#related-not-children) out of a set the +memory had it in, FIX-1451 into one the memory had it out of. Both were caught by reading the ticket +before acting on the memory, and only one of the two had a write pending. The habit is the control, +not the outcome. ## Who owns what @@ -129,11 +144,12 @@ it a proof rather than a surface; what it owns is [ER-20](BUSINESS-RULES.md), a rather than a matrix row, and **FIX-1385 provides the surface ER-20 stands on without owning it**. [FIX-817 and FIX-1415](SPEC.md#related-not-children) consume ER-3 from outside the set. -**The matrix is drawn against the original five and has deliberately not been redrawn.** -[FIX-1381](#d6-extended) joined after it was drawn and **owns no cross-cutting rule here**. It has -no spec yet, so which ER rows it consumes is not yet knowable — and inventing cells for a child -whose scope nobody has written is a worse defect than a missing column. It gets its column when its -spec says what it touches. +**The matrix is drawn against the original five and has deliberately not been redrawn.** Both later +children — [FIX-1381](#d6-extended) and [FIX-1451](#d6-seven) — **own no cross-cutting rule here**. +FIX-1381 has no spec yet, so which ER rows it consumes is not yet knowable; FIX-1451 is a bug and +will never have a spec, so its column would be read off its PR rather than a scope document. +Inventing cells for either is a worse defect than a missing column. FIX-1381 gets its column when +its spec says what it touches; FIX-1451 gets one only if its fix turns out to touch a rule. ## Decided in review, recorded so no child reopens them @@ -151,18 +167,20 @@ spec says what it touches. | ER-3 means **(a)** a contrast with the declared tree — **not (b)** relocating ChannelFlow's post fence onto an org resource | A channel's `members` is already live, read on the refusal path in the session the post lands in; (b) would put one fact in two places and tax every post | ER-3 · ER-12 | | **Drain width is 1.** Work queues behind a busy seat; a busy seat never takes a second row concurrently | Closes FIX-1408's *auto-scale / busy-copy* wall on FIX-1430's evidence, the way [ER-15](BUSINESS-RULES.md) says such a wall closes: the lab ran at **both** widths and wrote the comparison out, rather than a number being picked. **Reversible** — the `MANAGER_QUEUE_DRAIN_WIDTH` knob stays, which is how the other case remains reachable | FIX-1430's `lab/README.md` · [Open](#open), one wall lighter | | **Documents are org-scoped for now** — a seat's document is simply a resource under that seat on the org, and there is no per-seat document slot in v1 | The owner, 2026-09-19, answering FIX-1394's second ratify fork **in his own words rather than by taking the recommendation**. Per-resource configurability is not refused, it is *deferred to a mechanism that does not exist yet*: **resource templates** — "used to define the resource, but isn't the resource itself". Until those exist there is nothing to configure against, which is why v1 gets the org scope and no slot. **This unblocks the ratify**, and with it the ship tickets ER-8 gates | [ER-2](BUSINESS-RULES.md) · FIX-1394's ratify | -| **The set is six.** [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) — the seat resource allowlist — is pulled in as a child | The owner, 2026-09-19 16:01, **"pull it in"**, answering whether to take it into the epic's tail or leave it in backlog. **His word, not an inference from an action.** It exists because he asked whether resource access inside workers and skills could be controlled yet and it cannot — `resourcesFromDocs` hard-codes `scope: "org"`, `workerConfigSchema` admits only `instructions` / `teamInstructions` / `seatSkills`, and no access gate exists in `workforce/src` or `orchestration/src`. No new ticket was needed: FIX-1381 already covers it as D-11 Ask 1's ship ticket. **It moves the wrap out**, knowingly | [D6](#d6-extended) · [the wrap](PLAN.md#wrap) | +| **The set grew to six.** [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) — the seat resource allowlist — is pulled in as a child | The owner, 2026-09-19 16:01, **"pull it in"**, answering whether to take it into the epic's tail or leave it in backlog. **His word, not an inference from an action.** It exists because he asked whether resource access inside workers and skills could be controlled yet and it cannot — `resourcesFromDocs` hard-codes `scope: "org"`, `workerConfigSchema` admits only `instructions` / `teamInstructions` / `seatSkills`, and no access gate exists in `workforce/src` or `orchestration/src`. No new ticket was needed: FIX-1381 already covers it as D-11 Ask 1's ship ticket. **It moves the wrap out**, knowingly | [D6](#d6-extended) · [the wrap](PLAN.md#wrap) | | **The epic treats the W3 ship fence as narrowed to per-PR file overlap**, from 2026-09-19 | **An operating default, not a ruling, and not the owner's.** He merged #1928 through the live fence; the epic reads that action — not any statement, because there was none — as licence to stop parking W4 ship PRs whose diffs do not touch an unlanded W3 child's files. Recorded so the epic's behaviour is predictable and attributable rather than improvised per PR. **[The re-gate](#er-14-re-gate) is still open and still the owner's**, and **he can countermand this in a sentence**, at which point the default reverts to the rule as written | [ER-14](BUSINESS-RULES.md)'s note · [the re-gate](#er-14-re-gate) | | **Two of FIX-1430's approved rules had their stated mechanism refuted by the code.** Both texts are corrected; both rules stand and are graded | **BR-5:** a row does run in its own session, with a parent bound at mint and none of the coordinator's transcript — but the parent is the **draining seat's** session, not the filer's, because the dispatch that mints the child happens at the seat, after the filing. **BR-11:** `adoptLapsedLease` (`packages/orchestration/src/task-board/task-entry.ts`) **renews** a lapsed lease so a successor *can* take the row back; `StaleTaskClaimError` fires only when a reclaim genuinely won. The substance — a lapsed row is really back in the queue — now grades the stronger true claim: the next drain takes it and runs it on the seat whose own file answers for the desk it was filed for | ER-4 · FIX-1430's `BUSINESS-RULES.md` and `lab/README.md` | +| **The set is seven — FIX-1451 was a child all along.** Not a new call; a correction to what this document said | The ticket, read directly rather than recalled: `parentId: FIX-1407`, a `## Parent` section naming this epic (*"soft encounter under package cohesion, not a ship-gate on FIX-1394"*), and Architect guidance reading *"ship-able honesty fix under FIX-1407; do not nest under FIX-1359"* — filed by the owner at 12:58 on 2026-09-19. The epic held a standing note to clear the parenting; **the note was wrong and the write was never made**. It is a `Bug`, so it [skips the spec](SPEC.md#the-seventh-child) and holds the wrap like any child | [D6](#d6-seventh) · [the wrap](PLAN.md#wrap) · ER-16 · ER-17 | The first four rows are round 1 (2026-09-18, five reviewers on [#1905](https://github.com/fixpoint-labs/flow-state-dev/pull/1905)); the rest are 2026-09-19, raised up from a child rather than decided locally ([ER-15](BUSINESS-RULES.md) working). **The drain-width and refuted-mechanism rows are a child's PR correcting the epic** — the same mechanism as the -cross-spec pass above them, pointed at a fact rather than a scope. **[The fence -default](#fence-default) is the one row that is neither**: it is how the epic behaves while a -question with the owner stays open, and it decides nothing that question will decide. None of them -reopens an approved decision. +cross-spec pass above them, pointed at a fact rather than a scope. **Two rows are neither.** [The +fence default](#fence-default) is how the epic behaves while a question with the owner stays open, +and it decides nothing that question will decide. [The seventh child](#d6-seven) is not a decision +at all — it is the tracker correcting this document. None of these rows reopens an approved +decision. **Two stay falsifiable.** ER-3's reading flips on evidence that a channel's `members` is *not* current on the refusal path; the ER-15 split flips if those walls turn out to carry package content @@ -179,29 +197,20 @@ answer: are the package format and the board's assign surface really two issues? **Both of FIX-1394's ratify forks are now closed** — authorship, and [documents](#documents-answer) on 2026-09-19 — so nothing here blocks a child any more. The set -question is closed and [extended](#d6-extended); the compose helper's public export closed with it -([D5](#d5)). What remains open is the FIX-1451 question, **a re-gate on ER-14**, and three of -FIX-1408's walls. **None of them blocks work today.** - -### Does FIX-1451 belong under this epic? - -*(Decides: the owner. Blocks: nothing today. Bites: the wrap.)* -Parented but never confirmed, so the tracker holds seven against a set of six — stated in full at -[the spec](SPEC.md#the-unconfirmed-child). **Recommendation: re-home it as related-not-child**, the -way FIX-817 and FIX-1415 went, because the exit gate does not depend on it. **What would change my -mind:** if skill-context honesty is part of what *package cohesion* is supposed to deliver, it is a -child and the wrap waits for it. **Note this is now the second thing added to the wrap's tail** — -FIX-1381 was [pulled in deliberately](#d6-extended), and confirming FIX-1451 would push the finish -out again, which is worth weighing as a pair rather than one at a time. +question is closed twice over: [extended](#d6-extended) to six, then [corrected](#d6-seventh) to +seven. The compose helper's public export closed with D5's card ([D5](#d5)). What remains open is +**a re-gate on ER-14** and three of FIX-1408's walls. **Neither blocks work today.** ### Does the W3 ship fence narrow to per-PR overlap? — a pending re-gate *(Decides: the owner, and only the owner — this document records him ratifying **the ship fence as written** at the objective gate, so narrowing it is a re-gate, not an epic call. Blocks: now -**FIX-1394's eventual ship PR**, **FIX-1381's eventual implementation** and the [wrap -term](PLAN.md#wrap); no longer #1928 — the bite **grew** with the set rather than shrinking as the -merges landed. The recommendation below is **the epic's**, and **unanswered**.)* +**FIX-1394's eventual ship PR**, **FIX-1381's eventual implementation**, **FIX-1451's fix if it +ships a package** — the changeset call on that bug has not been made, so this is a conditional, not +an assertion — and the [wrap term](PLAN.md#wrap); no longer #1928. The bite **grew** with the set +rather than shrinking as the merges landed. The recommendation below is **the epic's**, and +**unanswered**.)* **The facts.** #1928 publishes `@flow-state-dev/workforce` and carries a changeset, so it is a **ship PR** by [ER-14](BUSINESS-RULES.md)'s letter. FIX-1435 — a W3 child carrying an diff --git a/spec/_epics/work-routing-package-cohesion/PLAN.md b/spec/_epics/work-routing-package-cohesion/PLAN.md index 561b14884c..87b4aa6254 100644 --- a/spec/_epics/work-routing-package-cohesion/PLAN.md +++ b/spec/_epics/work-routing-package-cohesion/PLAN.md @@ -8,22 +8,23 @@ which is each issue's own plan. IDs cross-reference [DECISIONS.md](DECISIONS.md) ## The path -![The path: a ship-fence band, two W3 input lanes, the six lanes of the set against time with a now line at September 19, and two related lanes under a divider. The figure's aria-label carries every lane and bar.](figures/path.svg) +![The path: a ship-fence band, two W3 input lanes, the seven lanes of the set against time with a now line at September 19, and two related lanes under a divider. The figure's aria-label carries every lane and bar.](figures/path.svg) -Ten lanes: two W3 inputs, the **six** children of the set, and — under the divider — the two +Eleven lanes: two W3 inputs, the **seven** children of the set, and — under the divider — the two [related issues](SPEC.md#related-not-children) that still consume FIX-1405. Four lanes are merged; -FIX-1394 sits at its now-unblocked ratify, and **FIX-1381's lane opens at the now line with no bar -behind it**, which is what a child pulled in at 16:01 looks like. The lane the tracker holds and -this figure does not is [FIX-1451](SPEC.md#the-unconfirmed-child) — deliberately, until it is -confirmed a child or re-homed. +FIX-1394 sits at its now-unblocked ratify, and the last two open at the now line with no bar behind +them — **FIX-1381**, pulled in at 16:01, and **[FIX-1451](SPEC.md#the-seventh-child)**, which was a +child all along and is only now drawn. Two empty lanes at the right edge is what the wrap's tail +looks like. **The critical path ran boards → the proof, and all of it is behind us.** FIX-1385 merged on 2026-09-19 and FIX-1430's build followed it onto `main` the same day ([#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929)). **What stands between the epic and its gate is not work**: it is [ER-20's own run](SPEC.md#er-20-has-not-run), which needs a model credential this epic has never had — which is why the proof's lane ends in a gate box rather than at -its bar. **Behind the gate, the wrap now has a new tail:** FIX-1381 must also finish, and it has not -started. Package cohesion is *beside* the critical path; FIX-1381 is *after* it. +its bar. **Behind the gate, the wrap has a tail of three:** FIX-1394's ratify, FIX-1381's spec and +FIX-1451's fix. None is on the critical path — package cohesion runs *beside* it, and the two new +children come *after* it — but all three are terms of the wrap. ## What each issue entails @@ -34,12 +35,15 @@ started. Package cohesion is *beside* the critical path; FIX-1381 is *after* it. | **FIX-1385** channel boards | spec → POC → **impl, merged** | The ChannelFlow floor · ER-5 · ER-3, which it may not re-decide and reads only **if** the roster has landed | A channel holding `0..N` TaskCollections; two doors, one surface. Plus the PR-5 name check over its own diff (ER-19) | The proof · FIX-1430 — **released 2026-09-19** | Large · shipped in one PR, not four | | **FIX-1394** package cohesion | **POC matrix → ratify, now unblocked** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both landed** at `d8e4c99`, and the code wins over either spec (ER-18) · D4 | One format as a Markdown file, instructions and tools, two attachment modes, **not disk-only**, **documents org-scoped** (ER-2). Its POC also carries the evidence for **one** of FIX-1408's walls (ER-15) | Ship tickets, once the ratify completes (ER-8) — **not** the proof | Large | | **FIX-1381** seat resource allowlist | spec → impl · **spec being written**, no PR yet | **D-11's settled direction** (FIX-1380, Done): Ask 1 the thin allowlist, Ask 3 org `ro` automatic / `rw` by permission · today's `resourcesFromDocs` org hard-code and the `WorkerConfig` gap its invent-kill names | Thin seat/kind resource refs by `ro`/`rw` — the first control over which resources a worker or skill reaches | **The wrap** — it is now a term of it | Not sized · direction settled, spec outstanding | +| **FIX-1451** allowed-tools honesty · *a bug* | **no spec** → impl · dispatched on `fix/FIX-1451-allowed-tools-honesty`, no PR yet | FIX-1416's landed `tools:` fence at `d8e4c99` — the promise the bug is about (ER-18) | A seat's skill no longer promising a tool grant the loader does not make | **The wrap** — a term of it like any child. Explicitly **not** a ship-gate on FIX-1394 | Not sized · a bug, scoped by its PR rather than a spec | | **FIX-1430** manager-queue lab · *the proof* | spec → **impl, merged** — but the goal check is **NOT RUN** | ER-1 · **ER-2 as a fence, not an input** · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, queue columns as views. Plus the drain-width comparison ER-15 called for ([width 1](DECISIONS.md#drain-width)) | The epic's wrap, **once ER-20 runs** — merging did not release it | Medium · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) merged | **Size is a read, not an estimate.** Three of the original five are exploration tickets whose own scope is the thing being explored — FIX-1394 most of all, whose ship tickets are not in this set -yet. **FIX-1381 is deliberately unsized:** its direction is settled by D-11 but its spec is being -written, and a number before the spec would be a guess wearing an estimate's clothes. +yet. **The two newest children are deliberately unsized, for opposite reasons:** FIX-1381's +direction is settled by D-11 but its spec is still being written, and a number before the spec would +be a guess wearing an estimate's clothes; FIX-1451 is a bug and will never have a spec, so its scope +is whatever its PR turns out to be. ## Where it is @@ -62,26 +66,31 @@ the graph because they do not block, not because the code is missing. caller's `assignee` verbatim rather than resolving a seat ([ER-5](BUSINESS-RULES.md)). 4. **FIX-1385's implementation landed** ([#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922)) → the proof's build ran, and has since merged too. The set's only true blocking edge is - discharged, and **four of the six children are on `main`**. + discharged, and **four of the seven children are on `main`**. 5. **ER-20 runs on the real path** → the epic's *claim* is proved. It is blocked on an **environment**, not on work: every line of code it needs is merged, and - [the gate has not run](SPEC.md#er-20-has-not-run). It is no longer the last thing in front of - the wrap — [FIX-1381](#wrap) is the other — but it is the only one nobody can work around. + [the gate has not run](SPEC.md#er-20-has-not-run). It is not the last thing in front of the wrap + — [three open children](#wrap) are the rest — but it is the only one nobody can work around. 6. **W3's last implementing child merges** → the ship fence lifts (ER-14). That is **FIX-1435**, still `Todo` with no branch and no PR. Four W4 ship PRs have merged under the live fence, the last of them **#1928, by the owner himself** — which settles that PR without answering the rule. - What is still fenced is **FIX-1394's eventual ship PR**, **FIX-1381's eventual implementation** - and the [wrap](#wrap) — the fence's bite **grew** when the set did. Whether the rule + What is still fenced is **FIX-1394's eventual ship PR**, **FIX-1381's eventual implementation**, + **FIX-1451's fix if it ships a package** (that changeset call has not been made) and the + [wrap](#wrap) — the fence's bite **grew** when the set did. Whether the rule narrows is [a re-gate with the owner](DECISIONS.md#er-14-re-gate), not an epic call; how the epic behaves until he answers is [an operating default](DECISIONS.md#fence-default). 7. **FIX-1394's ratify completes** → ship tickets are cut (ER-8). **Both forks are now answered** — authorship, and [documents as org-scoped](DECISIONS.md#documents-answer) — so nothing external holds it. Those tickets are new children and will need this path redrawn. It does **not** unblock the proof ([the spec](SPEC.md#what-the-proof-consumes)). -8. **FIX-1381's spec lands, then its implementation** → the last wrap term other than ER-20 clears. - It starts with its direction already decided by D-11 (FIX-1380, Done), so the spec is describing - a settled shape rather than choosing one. Its implementation will be a **ship PR** and therefore - sits under [ER-14](BUSINESS-RULES.md). +8. **FIX-1381's spec lands, then its implementation** → one of the three remaining wrap terms + clears. It starts with its direction already decided by D-11 (FIX-1380, Done), so the spec is + describing a settled shape rather than choosing one. Its implementation will be a **ship PR** and + therefore sits under [ER-14](BUSINESS-RULES.md). +9. **FIX-1451's fix merges** → the last of the three clears. **No spec gate on this one** — it is a + `Bug`, so its PR is the review surface ([ER-17](BUSINESS-RULES.md)) and it goes straight to + implementation on `fix/FIX-1451-allowed-tools-honesty`. It is fenced by ER-14 only **if** it + ships a package; that call has not been made. ## Coordination seams to watch @@ -89,6 +98,7 @@ the graph because they do not block, not because the code is missing. |---|---|---| | The inventory surface | FIX-1405 and FIX-817 *(related, not a child)* | 1405 owns both layers; 817 reads them, and specs only after 1405's reader contract is approved (ER-23) | | The seat's `tools:` fence | FIX-1394 · FIX-1415, outside the set · FIX-1416, **landed** | Three edits on one fence, in that order (ER-23). None widens past seat `tools:`, and FIX-1416's merged code is the fence's actual shape, not its spec | +| The skill's tool promise | **FIX-1451** and FIX-1394 · FIX-1416, **landed** | The bug is that a skill's `allowed-tools` promises a grant the loader does not make, on the same surface FIX-1394's format scopes to *instructions and tools*. **Explicitly not a ship-gate on FIX-1394** — the owner filed it as a soft encounter, so 1394's ratify never waits on it. Both write against FIX-1416's merged fence at `d8e4c99`, not its spec (ER-18) | | The seat's prompt compose | FIX-1394 and FIX-1377 | Composes framework default → team → seat. No fourth layer; the team's text does not merge into the seat's | | The board's claim / assign surface | FIX-1385 (**merged**) and FIX-1430 (**#1929, merged**) | 1385 owned it and shipped it; the lab consumes what landed. Two of 1385's rules about that surface were read back by the lab and [corrected](DECISIONS.md#refuted-mechanisms) rather than worked around. What 1385 did **not** settle is named in [DECISIONS](DECISIONS.md#the-cross-flow-board-residue) and has no owner | | The same surface, again | FIX-1385 and FIX-1415 *(related, not a child)* | 1415's invite verbs land on the channel 1385 hangs a board on. Re-homing moved the wrap, not the collision | @@ -118,16 +128,18 @@ FIX-1410, FIX-1373, FIX-957. ## Wrap -When **[ER-20](BUSINESS-RULES.md) holds** and every remaining child is terminal — **now six -children, not five** ([D6 as amended](DECISIONS.md#d6-extended)). +When **[ER-20](BUSINESS-RULES.md) holds** and every remaining child is terminal — **now seven +children, not five** ([amended](DECISIONS.md#d6-extended), then [corrected](DECISIONS.md#d6-seventh)). **Terminal is read from GitHub, not from Linear.** A child is terminal when its **implementation PR is merged or closed**, or when this epic-spec records its work as **done by decision rather than as code** — which is FIX-1408, and the set table is where that is recorded. Linear is the mirror; a merge is what makes code shipped. **Four children are terminal: FIX-1408 by decision, FIX-1385, -FIX-1405 and FIX-1430 by merge.** Two are not: FIX-1394, at its now-unblocked ratify, and -**FIX-1381, which has no spec yet**. +FIX-1405 and FIX-1430 by merge.** Three are not: FIX-1394 at its now-unblocked ratify, **FIX-1381, +which has no spec yet**, and **FIX-1451, which has no PR yet**. Note that FIX-1451 needs no spec to +become terminal — a bug's route runs straight to its implementation PR +([ER-17](BUSINESS-RULES.md)) — so a missing spec there is not a missing step. **ER-20 has not run, so the gate has not been met.** #1929 merged, which made FIX-1430 terminal and left the gate exactly as unrun as before — [the row reads `NOT RUN`](SPEC.md#er-20-has-not-run) and @@ -136,16 +148,17 @@ this set is the clearest possible case of the difference: everything the first c `main`, and nobody has watched it work. Somebody running that check with a model credential, and the verdict log saying PASS, is what meets the gate. -**And meeting the gate is no longer the same moment as wrapping.** [D6](DECISIONS.md#d6-extended) -bought exactly that alignment at the objective gate, and the owner deliberately spent part of it on -2026-09-19 by pulling **FIX-1381** into the set — a child with no spec and no implementation. So the -wrap now needs **ER-20 passing *and* FIX-1381 terminal**, and the second of those has not started. -**This was not a side effect:** the consequence was stated before the call, and the call was his. - -**Which children count is still partly open.** The tracker holds a **seventh**, -[FIX-1451](SPEC.md#the-unconfirmed-child), parented but never confirmed, and this term reads *every -remaining child*. Until it is confirmed in or re-homed out, the wrap has an input nothing here -sequences — and confirming it would extend the tail a second time. +**And meeting the gate is no longer the same moment as wrapping.** [D6](DECISIONS.md#d6) bought +exactly that alignment at the objective gate, and the set has since grown twice — the owner +deliberately pulled **FIX-1381** in at 16:01 with the cost stated to him, and **FIX-1451** turned +out to have been a child since 12:58, which no call made and no call can undo. So the wrap now needs +**ER-20 passing *and* three open children terminal**. Only one of those two growths was a choice. + +**Which children count is no longer open.** The tracker and this document agree on seven +([ER-16](BUSINESS-RULES.md)), and there is no unconfirmed remainder. That agreement was reached by +reading the tickets rather than trusting this document, which is the only reason it is worth +stating — [the epic had FIX-1451 wrong](DECISIONS.md#d6-seventh), and a wrap computed from the +wrong membership is a wrap that closes over unbuilt work. **Three things must be named out at the wrap, not confirmed.** None has an owner inside W4, and each would otherwise close as a pass nobody ran: diff --git a/spec/_epics/work-routing-package-cohesion/SPEC.md b/spec/_epics/work-routing-package-cohesion/SPEC.md index 9488be0c68..1d5a5884c4 100644 --- a/spec/_epics/work-routing-package-cohesion/SPEC.md +++ b/spec/_epics/work-routing-package-cohesion/SPEC.md @@ -41,16 +41,17 @@ document:** read the spec branch's head, or the Linear document on the issue. | FIX-1385 | A channel holding `0..N` TaskCollections; channel actions and `taskTools` as two doors on one surface. Plus the PR-5 name check over its own diff ([ER-19](BUSINESS-RULES.md)) | **The exit gate's surface** — the board in "channel/org board → seat runs" | **Done** · 2026-09-19 · [#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922) **merged** — one full-scope PR, not the spec's four ([the seam it discharged](PLAN.md#coordination-seams-to-watch)) | | FIX-1430 | Manager-queue lab: a coordinator seat owns a channel board, assigns to linked seats, shows queue state as **views** over existing task status | The **proof** of the exit gate, and the owner of [ER-20](BUSINESS-RULES.md) | **Done** · 2026-09-19 · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **merged** — but **the gate it exists to run [has not run](#er-20-has-not-run)** | | [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) | **Seat resource allowlist** — thin seat/kind refs by `ro`/`rw`. The ship ticket for D-11 Ask 1 | Nothing today can control which resources a worker or skill reaches: `resourcesFromDocs` hard-codes `scope: "org"` and `workerConfigSchema` admits only `instructions`, `teamInstructions`, `seatSkills`. **Pulled in by the owner** ([below](#the-sixth-child-pulled-in)) | **Spec in progress** · `Backlog`, moving to a spec-writing status · spec being written on `spec/FIX-1381` · **no spec PR yet** | +| [FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451) | **Skill `allowed-tools` promises a grant it does not make** — the honesty fix on the seat's tool surface | A seat's package is half of this epic's title, and a promise the loader does not keep is the package lying about what a seat can reach. Parented by the owner as a *soft encounter under package cohesion*, explicitly **not a ship-gate on FIX-1394** ([below](#the-seventh-child)) | **Implementation in progress** · a `Bug`, so [**no spec**](#the-seventh-child) · `Backlog`, moving to in-progress · dispatched on `fix/FIX-1451-allowed-tools-honesty` · **no PR yet** | -**Four of six children are complete; two are open.** FIX-1408 done by decision, and FIX-1385, -FIX-1405 and FIX-1430 all merged to `main` on 2026-09-19. **FIX-1394** is unblocked at its ratify -now that [the documents question is answered](DECISIONS.md#documents-answer). **FIX-1381** was -pulled into the set at 16:01 and has neither a spec nor an implementation — the least-advanced child -by some distance. +**Four of seven children are complete; three are open.** FIX-1408 done by decision, and FIX-1385, +FIX-1405 and FIX-1430 all merged to `main` on 2026-09-19. Open: **FIX-1394** at its now-unblocked +ratify, **FIX-1381** at its spec, and **FIX-1451** at its implementation. The last two both joined +the set today — one pulled in, one confirmed as having been a child all along. -**Read that as four of six, not as the epic being done.** Two separate things are true and neither -implies the other: most of the set is merged, and **the thing the set exists to prove has never been -observed**. Nor is the finish line where it was this morning — [the set grew](#the-sixth-child-pulled-in). +**Read that as four of seven, not as the epic being done.** Two separate things are true and neither +implies the other: most of what is *built* is merged, and **the thing the set exists to prove has +never been observed**. Nor is the finish line where it was this morning — the set grew +[once](#the-sixth-child-pulled-in), then [again](#the-seventh-child). **[ER-20](BUSINESS-RULES.md) is OPEN. Its row reads `NOT RUN`.** The acceptance goal calls a real @@ -61,8 +62,9 @@ is enforced against ER-20 as written:** W4 does not wrap until that row says PAS [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **has since merged, and it did not change this.** A merged proof whose gate never ran is exactly the *claiming routing works rather than having seen it* that ER-20 exists to prevent, and it is now the literal state of the set. What -stands between W4 and its wrap is **a machine with an inference key, and [FIX-1381](#the-sixth-child-pulled-in)** -— the second is work that can be done, the first is not. +stands between W4 and its wrap is **a machine with an inference key, and the three open children** — +FIX-1394's ratify, [FIX-1381](#the-sixth-child-pulled-in)'s spec and +[FIX-1451](#the-seventh-child)'s fix. Those three are work that can be done. The key is not. **The sixth child, and what it cost.** [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) @@ -92,15 +94,27 @@ question from which resources a seat may reach. [FIX-1442](https://linear.app/fixpoint-labs/issue/FIX-1442), the org-identity security pass, is adjacent. Neither is folded in here. + -**And FIX-1451 is still unconfirmed** — separately from the above. -[FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451), a `Bug` (*skill `allowed-tools` -promises a grant it does not make*), was parented on 2026-09-19 but never confirmed as a child, so -the tracker now holds **seven** against a set of six. Real work nobody disputes; the question is -only whether this epic's wrap waits for it. **Not re-parented here** — re-parenting is a deliberate -act with its own evidence, never a side effect of a refresh. Still -[an open question](DECISIONS.md#open). +**The seventh child: FIX-1451 was never unconfirmed — the epic's note was wrong.** +[FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451) (*skill `allowed-tools` promises a grant +it does not make*) carries `parentId: FIX-1407`, a `## Parent` section naming this epic — *"soft +encounter under package cohesion, not a ship-gate on FIX-1394"* — and Architect guidance reading +*"ship-able honesty fix under FIX-1407; do not nest under FIX-1359"*. Filed by the owner at 12:58 on +2026-09-19. **The parenting is deliberate and documented.** The epic carried a standing note saying +to clear it; that note was wrong and the write was never made, which is the only reason the mistake +cost nothing. + +**It is a `Bug`, so it skips the spec** and enters at implementation with its PR as the review +surface ([ER-17](BUSINESS-RULES.md)) — there is no spec gate to wait for. Dispatched on +`fix/FIX-1451-allowed-tools-honesty`. **It holds the wrap like any child**, so the wrap's tail is +now three open children plus ER-20. + +**Worth naming, because it is this epic's own recurring defect in miniature:** a remembered decision +about set membership, contradicted by the ticket, caught only by reading the ticket before acting on +the memory. That is twice in one day — [FIX-817](#related-not-children) went the other way, out of a +set the memory had it in. **Related, not children.** [FIX-817](https://linear.app/fixpoint-labs/issue/FIX-817) (catalog @@ -127,6 +141,8 @@ flowchart LR BD -->|"the exit gate's surface"| LAB["FIX-1430 · manager-queue lab · the proof"] PK -.->|"a fence on what the lab may author — not an input"| LAB D11["FIX-1380 · D-11 · done"] -.->|"the settled direction: thin allowlist, org ro/rw"| ACL["FIX-1381 · seat resource allowlist"] + T416 -.->|"the fence whose promise the bug breaks — not a wait"| HON["FIX-1451 · allowed-tools honesty · bug"] + HON -.->|"soft encounter — explicitly not a ship-gate"| PK subgraph REL ["related, not in the set — consume it, do not hold its wrap"] CAT["FIX-817 · catalog manifests"] ADM["FIX-1415 · channel-admin"] @@ -198,11 +214,15 @@ ratify is unblocked. **One thing is between W4 and the gate, and it is not work:** **a machine with an inference key.** That is the whole of what [ER-20](#er-20-has-not-run) waits for — every line of code it needs is -merged. Behind the gate, the wrap also now waits on **FIX-1381**, which has yet to be specced -([above](#the-sixth-child-pulled-in)). +merged. Behind the gate, the wrap waits on **three open children** — FIX-1394's ratify, +[FIX-1381](#the-sixth-child-pulled-in)'s spec and [FIX-1451](#the-seventh-child)'s fix. All three +can start today; none of them can substitute for the key. -**Open with you, blocking nothing today:** **whether FIX-1451 belongs under this epic** -([above](#the-unconfirmed-child)), and a **[pending re-gate on +**Settled since the last refresh, and it was the epic that had it wrong:** **FIX-1451 is a child**, +parented deliberately and documented as such on the ticket ([above](#the-seventh-child)) — the +epic's standing note to clear it was mistaken, and no write was ever made on it. + +**Open with you, blocking nothing today:** a **[pending re-gate on ER-14](DECISIONS.md#er-14-re-gate)** — you merged #1928 through the fence, which settles that PR but says nothing about the rule; narrowing it reopens what you ratified as item 3 above, so only you can do it. **With the epic:** FIX-1408's three remaining session-policy walls diff --git a/spec/_epics/work-routing-package-cohesion/figures/path.svg b/spec/_epics/work-routing-package-cohesion/figures/path.svg index 84456deeaa..fe5520efce 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/path.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/path.svg @@ -1,5 +1,5 @@ - + From 26848c326efec4d0aa907a481646c6a10cb81bca Mon Sep 17 00:00:00 2001 From: Claude Date: Sat, 19 Sep 2026 16:44:10 +0000 Subject: [PATCH 17/21] =?UTF-8?q?spec(epic):=20FIX-1381=20#1935=20and=20FI?= =?UTF-8?q?X-1451=20#1936=20=E2=80=94=20both=20placeholders=20filled=20(FI?= =?UTF-8?q?X-1407)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The last two lanes of the set had no PR when they were written. They do now: - FIX-1381 — spec PR #1935 open on spec/FIX-1381, in spec review. The spec gate is with the owner and unanswered; nothing on it is approved. - FIX-1451 — fix PR #1936 open on fix/FIX-1451-allowed-tools-honesty, 10 files. A Bug, so no spec (ER-17); its PR is the review surface. Set table, path.svg's last two lanes and its aria-label, plus the sentences that restated "no PR yet" in PLAN.md, DECISIONS.md and ER-17. Neither PR is merged, so neither child is terminal: the wrap's tail is still three open children plus ER-20. ER-20 is untouched and still NOT RUN — two more children having PRs does not move the exit gate, which still needs a machine with an inference key. ER-14's row is byte-identical, fifth refresh running. Counts re-derived from the table: 7 rows, 4 Done, 3 open. The heading read "6 issues" since the sixth child landed; corrected to 7. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01PxDE76gD3BxXnPPjT7BT1T --- .../BUSINESS-RULES.md | 2 +- .../DECISIONS.md | 6 +++--- .../work-routing-package-cohesion/PLAN.md | 16 +++++++++------- .../work-routing-package-cohesion/SPEC.md | 19 ++++++++++--------- .../figures/path.svg | 12 +++++++----- 5 files changed, 30 insertions(+), 25 deletions(-) diff --git a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md index 25c5da78a9..daf1804e4f 100644 --- a/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md +++ b/spec/_epics/work-routing-package-cohesion/BUSINESS-RULES.md @@ -71,7 +71,7 @@ with the owner, where the recommendation is the epic's and unanswered. |---|---|---| | ER-15 | A cross-cutting question is commented **up** on the epic PR, never decided locally. FIX-1408's unclosed walls are **epic-owned**, and their current standing is in [Open](DECISIONS.md#open) | A local answer is a second authority. A wall closes on evidence a child produced and folded up here — never on a child's own say-so | | ER-16 | A child's Linear state is mirrored the moment it changes. **Holding as of 2026-09-19**, the egress outage over and the queued backlog written | The epic wake derives blocked-by from Linear, so stale state silently mis-orders the set. The wrap term is **not** re-coupled to Linear by this — it stays [derived from GitHub](PLAN.md#terminal), because what makes code shipped is a merge. What Linear *does* decide is which children exist, and it decided it twice on 2026-09-19: the set is **seven** — FIX-1381 pulled in by the owner ([D6](DECISIONS.md#d6-extended)), and [FIX-1451](SPEC.md#the-seventh-child) confirmed as having been parented all along ([the correction](DECISIONS.md#d6-seventh)). **No unconfirmed remainder.** The second of those is the rule earning its place: the tracker was right and this spec was wrong | -| ER-17 | Every child routes to *spec* by default; only a `Bug` label re-routes | Fail-closed. The original five carry `Design` + `Feature`, and **FIX-1381 is routing to spec** like any other child — its spec is being written now. **[FIX-1451](SPEC.md#the-seventh-child) is the first child to take the other branch:** labelled `Bug`, so it skips the spec and enters at implementation with its PR as the review surface. There is no spec gate to wait for on it, and its absence is not a gap | +| ER-17 | Every child routes to *spec* by default; only a `Bug` label re-routes | Fail-closed. The original five carry `Design` + `Feature`, and **FIX-1381 is routing to spec** like any other child — its spec is in review on #1935. **[FIX-1451](SPEC.md#the-seventh-child) is the first child to take the other branch:** labelled `Bug`, so it skips the spec and enters at implementation with its PR as the review surface. There is no spec gate to wait for on it, and its absence is not a gap | | ER-18 | A child depending on FIX-1377 or FIX-1416 writes against the **landed code** at `d8e4c99` and names it. Where a W3 spec and that code disagree, **the code wins** | **FIX-1416 shipped stricter than its spec**, which said a seat's own `blocks/` folder is callable without being listed anywhere. `resolveDeclaredTools` (`packages/workforce/src/hire.ts`) promotes only names declared in `tools:`, so a child carrying the spec's sentence forward builds against behaviour that does not exist | | ER-19 | The project's **PR-5** — a new surface uses the settled Layer 2 name, never the superseded one — is met here as **a check over this set's own diff**, owned by **FIX-1385**. It is **not** a repo-wide rename | The repo-wide pass waits on the project's vocabulary lock (**PD-4** — running it early pays the rename twice), so it stays the project's and has **no owner inside W4**: named out here rather than dropped, so nobody wraps this epic believing it ran | | ER-23 | **A shared contract's spec is approved before the next child specs against it.** FIX-1405's reader contract — **`readDeclaredRoster(root)`** on `@flow-state-dev/workforce/loader`, collecting rather than throwing, joining the layers on **`id` alone** — closes before FIX-817 starts; the `tools:` fence settles across FIX-1416 → FIX-1394 → FIX-1415 in that order | Two issues designing one discovery API at once is how you ship two of them. It outlives FIX-817 leaving the set: the dependency is on the contract, not the parent | diff --git a/spec/_epics/work-routing-package-cohesion/DECISIONS.md b/spec/_epics/work-routing-package-cohesion/DECISIONS.md index 9ac3349ee4..d103574714 100644 --- a/spec/_epics/work-routing-package-cohesion/DECISIONS.md +++ b/spec/_epics/work-routing-package-cohesion/DECISIONS.md @@ -146,10 +146,10 @@ rather than a matrix row, and **FIX-1385 provides the surface ER-20 stands on wi **The matrix is drawn against the original five and has deliberately not been redrawn.** Both later children — [FIX-1381](#d6-extended) and [FIX-1451](#d6-seven) — **own no cross-cutting rule here**. -FIX-1381 has no spec yet, so which ER rows it consumes is not yet knowable; FIX-1451 is a bug and -will never have a spec, so its column would be read off its PR rather than a scope document. +FIX-1381's spec is in review and unapproved, so which ER rows it consumes is not yet settled; +FIX-1451 is a bug and will never have a spec, so its column would be read off its PR rather than a scope document. Inventing cells for either is a worse defect than a missing column. FIX-1381 gets its column when -its spec says what it touches; FIX-1451 gets one only if its fix turns out to touch a rule. +its approved spec says what it touches; FIX-1451 gets one only if its fix turns out to touch a rule. ## Decided in review, recorded so no child reopens them diff --git a/spec/_epics/work-routing-package-cohesion/PLAN.md b/spec/_epics/work-routing-package-cohesion/PLAN.md index 87b4aa6254..656451a395 100644 --- a/spec/_epics/work-routing-package-cohesion/PLAN.md +++ b/spec/_epics/work-routing-package-cohesion/PLAN.md @@ -34,16 +34,16 @@ children come *after* it — but all three are terms of the wrap. | **FIX-1405** runtime inventory | spec → POC → **impl, merged** | The three existing readers · hire · `channelInstances` · open sessions (D5) | Two layers: a declared roster composed at read time (replacing both labs' private `LabRoster`), and the live org resource | FIX-817, on its approved reader contract (ER-23) · FIX-1415 · `team.*` wildcards, later. **Not FIX-1385** | Medium · all three PRs on `main` — PR-A [#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920), S4 [#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923), S5–S7 [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) | | **FIX-1385** channel boards | spec → POC → **impl, merged** | The ChannelFlow floor · ER-5 · ER-3, which it may not re-decide and reads only **if** the roster has landed | A channel holding `0..N` TaskCollections; two doors, one surface. Plus the PR-5 name check over its own diff (ER-19) | The proof · FIX-1430 — **released 2026-09-19** | Large · shipped in one PR, not four | | **FIX-1394** package cohesion | **POC matrix → ratify, now unblocked** → ship tickets | FIX-1377's team layer · FIX-1416's `tools:` fence — **both landed** at `d8e4c99`, and the code wins over either spec (ER-18) · D4 | One format as a Markdown file, instructions and tools, two attachment modes, **not disk-only**, **documents org-scoped** (ER-2). Its POC also carries the evidence for **one** of FIX-1408's walls (ER-15) | Ship tickets, once the ratify completes (ER-8) — **not** the proof | Large | -| **FIX-1381** seat resource allowlist | spec → impl · **spec being written**, no PR yet | **D-11's settled direction** (FIX-1380, Done): Ask 1 the thin allowlist, Ask 3 org `ro` automatic / `rw` by permission · today's `resourcesFromDocs` org hard-code and the `WorkerConfig` gap its invent-kill names | Thin seat/kind resource refs by `ro`/`rw` — the first control over which resources a worker or skill reaches | **The wrap** — it is now a term of it | Not sized · direction settled, spec outstanding | -| **FIX-1451** allowed-tools honesty · *a bug* | **no spec** → impl · dispatched on `fix/FIX-1451-allowed-tools-honesty`, no PR yet | FIX-1416's landed `tools:` fence at `d8e4c99` — the promise the bug is about (ER-18) | A seat's skill no longer promising a tool grant the loader does not make | **The wrap** — a term of it like any child. Explicitly **not** a ship-gate on FIX-1394 | Not sized · a bug, scoped by its PR rather than a spec | +| **FIX-1381** seat resource allowlist | spec → impl · **spec PR [#1935](https://github.com/fixpoint-labs/flow-state-dev/pull/1935) open**, in review, gate unanswered | **D-11's settled direction** (FIX-1380, Done): Ask 1 the thin allowlist, Ask 3 org `ro` automatic / `rw` by permission · today's `resourcesFromDocs` org hard-code and the `WorkerConfig` gap its invent-kill names | Thin seat/kind resource refs by `ro`/`rw` — the first control over which resources a worker or skill reaches | **The wrap** — it is now a term of it | Not sized · direction settled, spec in review | +| **FIX-1451** allowed-tools honesty · *a bug* | **no spec** → impl · fix PR [#1936](https://github.com/fixpoint-labs/flow-state-dev/pull/1936) open on `fix/FIX-1451-allowed-tools-honesty` | FIX-1416's landed `tools:` fence at `d8e4c99` — the promise the bug is about (ER-18) | A seat's skill no longer promising a tool grant the loader does not make | **The wrap** — a term of it like any child. Explicitly **not** a ship-gate on FIX-1394 | Not sized · a bug, scoped by its PR rather than a spec | | **FIX-1430** manager-queue lab · *the proof* | spec → **impl, merged** — but the goal check is **NOT RUN** | ER-1 · **ER-2 as a fence, not an input** · ER-4 · ER-5 — consumes four, owns none | A coordinator seat assigning over a channel board to linked seats, queue columns as views. Plus the drain-width comparison ER-15 called for ([width 1](DECISIONS.md#drain-width)) | The epic's wrap, **once ER-20 runs** — merging did not release it | Medium · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) merged | **Size is a read, not an estimate.** Three of the original five are exploration tickets whose own scope is the thing being explored — FIX-1394 most of all, whose ship tickets are not in this set yet. **The two newest children are deliberately unsized, for opposite reasons:** FIX-1381's -direction is settled by D-11 but its spec is still being written, and a number before the spec would -be a guess wearing an estimate's clothes; FIX-1451 is a bug and will never have a spec, so its scope -is whatever its PR turns out to be. +direction is settled by D-11 but its spec is still in review, and a number before that spec is +approved would be a guess wearing an estimate's clothes; FIX-1451 is a bug and will never have a +spec, so its scope is whatever its PR turns out to be. ## Where it is @@ -137,8 +137,10 @@ is merged or closed**, or when this epic-spec records its work as **done by deci code** — which is FIX-1408, and the set table is where that is recorded. Linear is the mirror; a merge is what makes code shipped. **Four children are terminal: FIX-1408 by decision, FIX-1385, FIX-1405 and FIX-1430 by merge.** Three are not: FIX-1394 at its now-unblocked ratify, **FIX-1381, -which has no spec yet**, and **FIX-1451, which has no PR yet**. Note that FIX-1451 needs no spec to -become terminal — a bug's route runs straight to its implementation PR +whose spec PR [#1935](https://github.com/fixpoint-labs/flow-state-dev/pull/1935) is open and +unapproved**, and **FIX-1451, whose fix PR +[#1936](https://github.com/fixpoint-labs/flow-state-dev/pull/1936) is open and unmerged**. Note +that FIX-1451 needs no spec to become terminal — a bug's route runs straight to its implementation PR ([ER-17](BUSINESS-RULES.md)) — so a missing spec there is not a missing step. **ER-20 has not run, so the gate has not been met.** #1929 merged, which made FIX-1430 terminal and diff --git a/spec/_epics/work-routing-package-cohesion/SPEC.md b/spec/_epics/work-routing-package-cohesion/SPEC.md index 1d5a5884c4..6745e2f19d 100644 --- a/spec/_epics/work-routing-package-cohesion/SPEC.md +++ b/spec/_epics/work-routing-package-cohesion/SPEC.md @@ -2,7 +2,7 @@ **Spec** · [Decisions](DECISIONS.md) · [Rules](BUSINESS-RULES.md) · [Plan](PLAN.md) -Epic · 6 issues · Workforce: Layer 2 Abstraction · Goal 1 — Workforce multi-seat real usage / +Epic · 7 issues · Workforce: Layer 2 Abstraction · Goal 1 — Workforce multi-seat real usage / architecture cohesion ## Four teams, before and after @@ -40,8 +40,8 @@ document:** read the spec branch's head, or the Linear document on the issue. | FIX-1405 | Inventory in two layers: a declared roster composed from the existing readers, and the live org resource ChannelFlow updates | The epic's third promise in its own right — *which seats and channels exist* — and what `team.*` addressing calls later ([ER-7](BUSINESS-RULES.md)). It does **not** gate the boards | **Done** · 2026-09-19 · PR-A [#1920](https://github.com/fixpoint-labs/flow-state-dev/pull/1920), S4 [#1923](https://github.com/fixpoint-labs/flow-state-dev/pull/1923) and S5–S7 [#1928](https://github.com/fixpoint-labs/flow-state-dev/pull/1928) all **merged** — terminal by merge | | FIX-1385 | A channel holding `0..N` TaskCollections; channel actions and `taskTools` as two doors on one surface. Plus the PR-5 name check over its own diff ([ER-19](BUSINESS-RULES.md)) | **The exit gate's surface** — the board in "channel/org board → seat runs" | **Done** · 2026-09-19 · [#1922](https://github.com/fixpoint-labs/flow-state-dev/pull/1922) **merged** — one full-scope PR, not the spec's four ([the seam it discharged](PLAN.md#coordination-seams-to-watch)) | | FIX-1430 | Manager-queue lab: a coordinator seat owns a channel board, assigns to linked seats, shows queue state as **views** over existing task status | The **proof** of the exit gate, and the owner of [ER-20](BUSINESS-RULES.md) | **Done** · 2026-09-19 · [#1929](https://github.com/fixpoint-labs/flow-state-dev/pull/1929) **merged** — but **the gate it exists to run [has not run](#er-20-has-not-run)** | -| [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) | **Seat resource allowlist** — thin seat/kind refs by `ro`/`rw`. The ship ticket for D-11 Ask 1 | Nothing today can control which resources a worker or skill reaches: `resourcesFromDocs` hard-codes `scope: "org"` and `workerConfigSchema` admits only `instructions`, `teamInstructions`, `seatSkills`. **Pulled in by the owner** ([below](#the-sixth-child-pulled-in)) | **Spec in progress** · `Backlog`, moving to a spec-writing status · spec being written on `spec/FIX-1381` · **no spec PR yet** | -| [FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451) | **Skill `allowed-tools` promises a grant it does not make** — the honesty fix on the seat's tool surface | A seat's package is half of this epic's title, and a promise the loader does not keep is the package lying about what a seat can reach. Parented by the owner as a *soft encounter under package cohesion*, explicitly **not a ship-gate on FIX-1394** ([below](#the-seventh-child)) | **Implementation in progress** · a `Bug`, so [**no spec**](#the-seventh-child) · `Backlog`, moving to in-progress · dispatched on `fix/FIX-1451-allowed-tools-honesty` · **no PR yet** | +| [FIX-1381](https://linear.app/fixpoint-labs/issue/FIX-1381) | **Seat resource allowlist** — thin seat/kind refs by `ro`/`rw`. The ship ticket for D-11 Ask 1 | Nothing today can control which resources a worker or skill reaches: `resourcesFromDocs` hard-codes `scope: "org"` and `workerConfigSchema` admits only `instructions`, `teamInstructions`, `seatSkills`. **Pulled in by the owner** ([below](#the-sixth-child-pulled-in)) | **In spec review** · spec PR [#1935](https://github.com/fixpoint-labs/flow-state-dev/pull/1935) open on `spec/FIX-1381` · **the spec gate is with the owner and unanswered — nothing here is approved** | +| [FIX-1451](https://linear.app/fixpoint-labs/issue/FIX-1451) | **Skill `allowed-tools` promises a grant it does not make** — the honesty fix on the seat's tool surface | A seat's package is half of this epic's title, and a promise the loader does not keep is the package lying about what a seat can reach. Parented by the owner as a *soft encounter under package cohesion*, explicitly **not a ship-gate on FIX-1394** ([below](#the-seventh-child)) | **In PR review** · a `Bug`, so [**no spec**](#the-seventh-child) — the PR is the review surface · fix PR [#1936](https://github.com/fixpoint-labs/flow-state-dev/pull/1936) open on `fix/FIX-1451-allowed-tools-honesty`, 10 files | **Four of seven children are complete; three are open.** FIX-1408 done by decision, and FIX-1385, FIX-1405 and FIX-1430 all merged to `main` on 2026-09-19. Open: **FIX-1394** at its now-unblocked @@ -78,8 +78,8 @@ or `orchestration/src`. It needed no new ticket: **FIX-1381 already covered it e ticket for D-11 Ask 1, its own invent-kill naming that `WorkerConfig` gap. **It moved the wrap out, and that was said before the call was made.** The [wrap](PLAN.md#wrap) now -needs FIX-1381 terminal **as well as** ER-20 passing, and FIX-1381 has no spec yet — so W4's finish -is further away today than it was this morning. The owner made that call with the consequence in +needs FIX-1381 terminal **as well as** ER-20 passing, and FIX-1381's spec is only now in review, +with no implementation — so W4's finish is further away today than it was this morning. The owner made that call with the consequence in front of him. **It enters cheaper than a cold child:** [D-11 (FIX-1380)](https://linear.app/fixpoint-labs/issue/FIX-1380) is **Done** and already decides its direction — Ask 1 the thin allowlist, Ask 3 org `ro` automatic and `rw` by permission — so only its spec is outstanding, not its shape. @@ -107,9 +107,10 @@ to clear it; that note was wrong and the write was never made, which is the only cost nothing. **It is a `Bug`, so it skips the spec** and enters at implementation with its PR as the review -surface ([ER-17](BUSINESS-RULES.md)) — there is no spec gate to wait for. Dispatched on +surface ([ER-17](BUSINESS-RULES.md)) — there is no spec gate to wait for. That surface now exists: +[#1936](https://github.com/fixpoint-labs/flow-state-dev/pull/1936) on `fix/FIX-1451-allowed-tools-honesty`. **It holds the wrap like any child**, so the wrap's tail is -now three open children plus ER-20. +still three open children plus ER-20. **Worth naming, because it is this epic's own recurring defect in miniature:** a remembered decision about set membership, contradicted by the ticket, caught only by reading the ticket before acting on @@ -215,8 +216,8 @@ ratify is unblocked. **One thing is between W4 and the gate, and it is not work:** **a machine with an inference key.** That is the whole of what [ER-20](#er-20-has-not-run) waits for — every line of code it needs is merged. Behind the gate, the wrap waits on **three open children** — FIX-1394's ratify, -[FIX-1381](#the-sixth-child-pulled-in)'s spec and [FIX-1451](#the-seventh-child)'s fix. All three -can start today; none of them can substitute for the key. +[FIX-1381](#the-sixth-child-pulled-in)'s spec and [FIX-1451](#the-seventh-child)'s fix. Two of the +three are now on open PRs; none of them can substitute for the key. **Settled since the last refresh, and it was the epic that had it wrong:** **FIX-1451 is a child**, parented deliberately and documented as such on the ticket ([above](#the-seventh-child)) — the diff --git a/spec/_epics/work-routing-package-cohesion/figures/path.svg b/spec/_epics/work-routing-package-cohesion/figures/path.svg index fe5520efce..a502fd9d34 100644 --- a/spec/_epics/work-routing-package-cohesion/figures/path.svg +++ b/spec/_epics/work-routing-package-cohesion/figures/path.svg @@ -1,5 +1,5 @@ + aria-label="The path: a ship-fence band under the axis saying FIX-1435 has no pull request and is still Todo, so the W3 ship fence has not lifted, and that four W4 ship PRs merged under the live fence, the last of them pull request 1928 by the owner himself. Below it, eleven lanes against time with a now line at September 19. Two input lanes from W3, TEAM.md and the tools fence, carry bars that reach the now line and read landed on main on September 19. Then the seven lanes of the set. FIX-1408 carries a done bar behind the now line, done by decision. FIX-1405 inventory opens with a short spec-approved bar at the now line and then carries one merged bar running to the end of its lane: done, with pull requests 1920, 1923 and 1928 all merged. FIX-1385 channel boards carries the same spec bar and a merged bar running to the end of its lane: Done, pull request 1922 merged, one full-scope PR rather than the four its plan named. FIX-1394 package cohesion carries an open-PR bar reading matrix pull request 1921, authorship decided, documents open — the documents half is now answered and this sentence in the figure is the state at the bar, not the state today. FIX-1430, the manager-queue lab, carries its spec bar, then a merged bar for pull request 1929, and then a dashed red gate box reading ER-20 NOT RUN — the code is merged and the exit gate's own run still needs a model credential this epic does not have. The last two lanes of the set each now open with a short PR-open bar at the now line and then a dashed not-started line, because each has a pull request and neither has finished. FIX-1381 seat resource allowlist reads spec pull request 1935 open, in review, not approved — the spec gate is still with the owner. FIX-1451 allowed-tools, a bug, reads fix pull request 1936 open, a bug so no spec — it needs no spec to finish, because a bug routes straight to its implementation PR. Neither bar is a merge, so both lanes plus FIX-1394's ratify are still the wrap's tail, alongside ER-20. Under a divider, two related lanes, catalog manifests and channel-admin, read related, not in the set: they consume FIX-1405 and hold no wrap, and catalog manifests waits for FIX-1405's approved spec. The critical path is drawn in the gutter through the boards and the lab only; the inventory lane sits beside it, not behind it.">