From bc749aa1fa456a3c4c5601020f8a754fc194371f Mon Sep 17 00:00:00 2001 From: Claude Date: Wed, 28 Oct 2026 01:38:08 +0000 Subject: [PATCH 1/5] spec(FIX-2242): closure QA plan for the FIX-2070 messaging epic One walk on one durable host where the epic's legs collide (a call across a crash, a midway question to a filer parked on a call, a ring beside it), one browser journey, the children's checks under a skip map, the gap sweep, and the removed message names as one rule in the CI sweep beside FIX-2134's. Co-Authored-By: Claude Opus 5.5 Claude-Session: https://claude.ai/code/session_01NczDiEtVsnByx4XbajAf5k --- specs/issues/FIX-2242/DECISIONS.md | 106 ++++++++++++ specs/issues/FIX-2242/SPEC.md | 151 ++++++++++++++++++ .../FIX-2242/figures/d1-legs-collide.svg | 100 ++++++++++++ .../FIX-2242/figures/d2-sweep-in-ci.svg | 100 ++++++++++++ 4 files changed, 457 insertions(+) create mode 100644 specs/issues/FIX-2242/DECISIONS.md create mode 100644 specs/issues/FIX-2242/SPEC.md create mode 100644 specs/issues/FIX-2242/figures/d1-legs-collide.svg create mode 100644 specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg diff --git a/specs/issues/FIX-2242/DECISIONS.md b/specs/issues/FIX-2242/DECISIONS.md new file mode 100644 index 0000000000..b3596568fe --- /dev/null +++ b/specs/issues/FIX-2242/DECISIONS.md @@ -0,0 +1,106 @@ +# FIX-2242 · Decisions + +[Spec](SPEC.md) · **Decisions** + +The plan's four rows are the [closure rule's](../../../docs/contributing/orchestration.md#the-closure-issue-every-epic-ends-in-qa) +and its legs are the [epic's](../../epics/FIX-2070/SPEC.md#the-goal-and-how-well-know-its-met). +Two calls are this spec's: how the walk puts the legs together, and where the sweep keeps running. + +## The tree + +```mermaid +flowchart TD + I["FIX-2242"] --> D1["D1 · one walk where the legs collide, one durable host"] + D1 -.->|"rejected · each child already walks its leg alone"| X1["the children's checks re-run side by side"] + I --> D2["D2 · the removed names are one rule in the CI sweep"] + D2 -.->|"rejected · a later PR brings a name back unseen"| X2["the sweep in one goal's run only"] +``` + +Solid edges are what you're signing. Dashed edges lost, and the label says why. + + +## D1 · One walk where the legs collide, on one durable host, not the children's checks side by side + +| | | +|---|---| +| **Instead of** | Re-running each child's legs (1 to 10c, j1) and calling the set proved when each is green | +| **Because** | Each child's check walks one leg on a host where nothing else happens. What none of them sees is the legs meeting: a call outstanding when the host dies between its row and its dispatch (a with c), a midway question reaching a filer whose request is parked on a call (e with a), a ring spending a typed message's budget while a call waits (d beside a). Those need the legs in one Lab, on one host, at once. The ring keeps its own typed message, so its budget can't be confused with the call's | +| **Locks in** | A real-model run with a kill and a restart, some minutes per attempt, repeated in full per finding. Each assertion is tagged `:`, so a red leg doesn't hide the others, and every attempt is printed | + +![D1: one walk where the legs collide, chosen, beside the children's checks re-run side by side; decided by collisions between legs, which a check of one leg can't produce](figures/d1-legs-collide.svg) + +It comes down to collisions: a check that walks one leg on a calm host can never produce one. + +**What would change my mind:** a host that can't be killed and restarted mid-call in a goal run. +Then a and c walk on their own, and e meets a parked call that never crosses a restart. + + +## D2 · The removed names are one rule in the CI sweep, beside FIX-2134's, not a check one goal runs + +| | | +|---|---| +| **Instead of** | `workers-message-each-other`'s own sweep (`run.mts`, `REMOVED`), which runs only when that goal runs, over `packages/*/src` | +| **Because** | The epic promises one entry and one tool. The likeliest way to break that after the wrap is a later PR bringing back `messageWorker`, the relay or a refusal under its old name, and only CI sees every PR. FIX-2087's engine already runs in CI with plants, and FIX-2134 added `PERSON_STOPS_REMOVED` beside its rows, so this is one more list in the same pass, not a second sweep. The goal's sweep then reads this list, so there is one copy | +| **Locks in** | A PR naming one of the twelve where the sweep reads (package source and tests, the pages a user reads; comments skipped) fails CI until it edits the list. `replyTo` is a word `core`'s dispatcher and tests still use for their own fields, so it is matched only in `packages/workforce`, `packages/shift-manager` and the pages, as FIX-2134 scoped `answerQuestion` | + +![D2: one rule in the CI sweep, chosen, beside the sweep in one goal's run; decided by a later PR bringing a removed name back](figures/d2-sweep-in-ci.svg) + +It comes down to a later PR: swept only by one goal, a returned `messageWorker` is seen at the next closure, if ever. + +**On `main` today the list is not yet green:** `packages/shift-manager/test/messages-view.test.tsx` +still renders a `takes_tasks_only` refusal, which FIX-2238 removed. The CI sweep reads tests, so +the PR adding the rule updates that fixture to a refusal that exists. + +## The plan in full + +| # | What runs | PASS means | +|---|---|---| +| 1 | Legs a to e and the sweep ([the signal](SPEC.md#the-signal)) | Each green; each control red where named | +| 2 | **j1** *uses Shift Manager*, in a browser as Alice after the walk: Helper A's conversation shows Researcher's answer once, as its own item, with its sender and the message it answers (`j1:own-item`); the Messages view lists the crashed call once, `delivered`, the parked call `answered`, and the ring's seventh `refused` (`j1:states`); each side's link opens its session (`j1:opens-side`). The other four teams' *after* is walked by leg a (builds talkers), e (files work), c and d (production) and the sweep (maintains); each mapping is printed | Each green | +| 3 | Every child's goal check with its controls, read from Linear at run time and printed: `workers-message-each-other` legs 1 to 10c with every control, FIX-2240's leg 9 included; `every-stop-is-one-inbox-item` j1 and j1b with FIX-2241's controls, and its other legs, since FIX-2241 changed its Lab. A check row 1 walks in full is skipped, printed beside the leg that covers it; one walked in part runs. #4021's two budget cases are CI tests | Each PASS beside its documented FAIL | +| 4 | The gap sweep: each [coordination seam](../../epics/FIX-2070/DECISIONS.md#coordination-seams) from both sides (on the send path, a `{ task }` target resolved before the limits, the call's wait after the dispatch, the resend through the wake; every field on the message row written by its owner and read by its reader; every worker flow, coordinators and the coder included, carrying `onMessageReceived`; the question sent through `sendMessage`'s server path, outside the budget; the Messages view's states, `undeliverable` and a waiting call included; FIX-2071's names untouched); [`messaging-workers.md`](../../../apps/docs/docs/workforce/messaging-workers.md) and the Workforce README followed as written, the page's call example run as a goal-local worker; each epic *not done if* state absent; the *Not doing* line on a session waiting on a user checked as a queue | Every seam two-sided | + +**Must be green first:** every other child merged with its goal's PASS and its controls' FAIL; +#4021 merged; the D2 rule in CI. **FIX-2240's leg 9 is not on `main`:** #4008 merged with its +model-free tests and without the goal leg its spec names (`9:crash-after-row` and the rest, with +`no-debt` and the other four controls). It lands before the run, as its own child of FIX-2070 +blocking this issue, since leg c reuses its faults and `no-debt`. + +## Decided, not asked + +- **Leg a's callee is an agent worker, not the coder.** The epic's figure says "writer calls + coder", but FIX-2238 D1 lets the coder hear only in its tasks' sessions. The coder is reached by + `{ task }`, which FIX-2160's leg 8 and FIX-2239's `10:task-call` cover in row 3. +- **Leg e asks through `askQuestion`, from an agent worker's task.** Both ask paths share one + point (`askMidway`); the coder's path is j1b, re-run in row 3. The walk then needs no Claude Code. +- **Leg c's crash uses FIX-2240's `exit-after-row` fault,** through `goals/lib/module-patch.mjs`, + one-shot on the call's row. +- **[FIX-2087 D1](../FIX-2087/DECISIONS.md#d1), inherited:** a structural miss on any attempt is a + finding; only a judgment miss (what the model chose to write) earns another attempt. +- **One new control, `call-holds-session`:** a parked call keeps its session's place, so a + question to the filer waits for the call to end. It is the one collision no child's control + breaks. A patch that never fired fails the build. +- **Today's `main` is `e2c190e`,** the commit before FIX-2238. Leg d can pass there, since + FIX-1989's hop limit already stopped a ring at 6; the epic names a, b, c and e. +- **No legacy leg:** stores are reset between the refactor epics, as the epic's seam table says. +- **Kitchen-sink's missing `workforce:received` renderer** is outside FIX-2070 and not graded. + +## Considered and dropped + +| Alternative | Why not | +|---|---| +| Re-run the children's checks and stop there | Each leg alone is proved by its own child; row 3 re-runs them. What none can show is the legs colliding (D1) | +| The ring in the same typed message as the call | Its budget and the call's would mix, and a red `d:seventh-refused` couldn't say whose message was seventh | +| A Claude Code coding run in the walk | The coder's ask path is j1b's; the collision is the filer's, whatever asks | +| The walk model-free | The epic's goal says real model; the CI tests are the model-free half | + +## How it got here + +- **Draft** — the epic's legs on one durable host, a call across the crash, a question to a + filer parked on a call; FIX-2134's shape and FIX-2087's D1 kept; the sweep added to the CI + engine beside `PERSON_STOPS_REMOVED`. Found on `main`: FIX-2240's leg 9 never landed, and a + shift-manager test still names `takes_tasks_only`. + +## Notes from review + +**Open: none of its own.** diff --git a/specs/issues/FIX-2242/SPEC.md b/specs/issues/FIX-2242/SPEC.md new file mode 100644 index 0000000000..6e1a3eb555 --- /dev/null +++ b/specs/issues/FIX-2242/SPEC.md @@ -0,0 +1,151 @@ +# FIX-2242 · Closure: workers talk by send and call, to any worker, task or conversation + +**Spec** · [Decisions](DECISIONS.md) + +## Five teams, before and after + +| A team that… | Today | After this closes | +|---|---|---| +| **decides whether the epic is done** | Each child green on its own commit; no check has a call, a crash, a midway question and a ring on one host at once | One report on one `main` commit, every check beside its control's FAIL | +| **uses Shift Manager, like Alice** | Each kind of message is shown in its own child's run | In a browser after the walk, each message shows once, in the session it is about, with its sender, what it answers and its state, a restart included | +| **builds a worker that talks to others** | A call is proved on a calm host | A call outstanding across a crash still returns its answer, once, to the request that asked | +| **builds a worker that files work** | Hears its task's midway question in a check with nothing else running | Hears it while its own call waits, steers, and the task goes on the same attempt | +| **runs Workforce in production, or maintains it** | The removed names are swept only when one goal runs | CI fails a later PR that brings `messageWorker` or the relay back ([D2](DECISIONS.md#d2)) | + +## The goal, and how we'll know it's met + +**On one `main` commit with every other child merged, in Shift Manager on a real model and a +durable host, a worker calls another and uses the answer in the request that asked, across a +crash between the log write and the dispatch; a message to a task lands in that task's session; +a task's midway question reaches its filer while the filer waits on a call; every message lands +once; a ring stops inside one typed message's budget; and none of the removed names is in source.** + +| Is it the right goal? | | +|---|---| +| **The real need** | [The epic's goal](../../epics/FIX-2070/SPEC.md#the-goal-and-how-well-know-its-met) on the assembled set: the gaps *between* children ([closure rule](../../../docs/contributing/orchestration.md#the-closure-issue-every-epic-ends-in-qa)) | +| **Smaller, and rejected** | "Re-run every child's legs." Each walks one leg on a calm host, so a call lost across a crash, or a question that can't reach a filer parked on a call, still passes ([D1](DECISIONS.md#d1)) | +| **Not done if** | The [epic's list](../../epics/FIX-2070/SPEC.md#the-goal-and-how-well-know-its-met), plus: a check ran on another commit · a structural miss was retried away ([FIX-2087 D1](../FIX-2087/DECISIONS.md#d1)) · a control was skipped or passed · a check graded through a child's helper · a finding is open | + +```mermaid +flowchart LR + M["one main commit · every child merged"] --> W["one Lab · durable host · real model"] + W --> A["a+c · a call, the host killed after its row"] + W --> E["e · a midway question while the filer's call waits"] + W --> B["b · a message to a task, open and finished"] + W --> D["d · a ring on its own typed message"] + A & E & B & D --> P["PASS · the epic's goal is met"] + K["controls · today's main · no-debt · send-again · no-question-message · call-holds-session"] -.->|"must FAIL"| A +``` + +The walk reads sessions and the log through public actions; each control breaks one leg. + +| How we verify | | +|---|---| +| **Goal check** | `goals/shift-manager/workers-talk-together/`, new, reusing `workers-message-each-other`'s cast and world · the model each worker's file names (`openai/gpt-5.4-mini`; the Shift Coordinator `anthropic/claude-haiku-5-5`) · run by the closure worker on one `main` commit · the report is the closure PR's body | +| **Signal** | Every `:` in [the signal table](#the-signal) passes | +| **Input** | DevTeam served by Shift Manager's command over a SQLite store file (`DEVTEAM_STORE`), killed and restarted on the same file; codenames and the brief's fact picked at run time | +| **Anti-game** | Assert on sessions' stored items, the message log and `getTask`, read as Alice; nothing seeded; sent messages counted from each turn's own tool outputs; a fault that never fired is a setup failure, not a grade | +| **Control that must fail** | Today's `main` before FIX-2238 (`e2c190e`): legs a, b, c and e, and the sweep · `no-debt` (FIX-2240's) **`c:lands-once`** · `send-again` (FIX-2239's) **`a:one-place`** · `task-to-conversation` (FIX-2160's) **`b:in-task-session`** · `no-question-message` (FIX-2241's) **`e:filer-hears`** · `no-hop-cap` (FIX-2238's) **`d:seventh-refused`** · `call-holds-session`, new: a parked call keeps its session's place **`e:heard-while-calling`** · the sweep's `--control` its plants | + +### The signal + +| Assertion | Passes when | +|---|---| +| `c:lands-once` | Helper A calls Researcher; the server kills itself after the call's row is written, before its dispatch; restarted without the fault, Researcher's session holds exactly one received item for the id, and the row is `delivered`, `deliveredAs` naming the resend | +| `a:answer-in-request` | Helper A's request that made the call, replayed by the durable host, reads Researcher's answer as the call's result, and its answer to Alice holds the codename, with no line from Alice in between. A killed request the host never replays is a finding, not a pass by waiting again | +| `a:one-place` | The log holds one row and one budget place for the call, crash and replay included | +| `a:no-relay` | Researcher's turn sends no message back, and Helper A's conversation gains no received item from it | +| `e:filer-hears` | Helper A files Helper B a task whose brief needs a fact only Helper A holds; Helper B asks it midway; the question arrives at Helper A's `onMessageReceived` in the session that filed the task | +| `e:heard-while-calling` | It arrives while Helper A's second call, into Researcher's conversation holding a stop for Alice, is parked: the received item is stored before the call resolves | +| `e:steer-releases` | Helper A's steer releases the hold; the task finishes on attempt 1; the call then resolves once, answered after Alice's answer, or kept and returned when Helper A waits again by call id | +| `b:in-task-session` | A message to Helper B's task while its turn runs lands in that task's session after the turn ends; one sent after the task finished is answered there from the task's history | +| `b:task-unchanged` | The row's status, output, attempts and session are the same before and after both | +| `d:per-request` | Started on its own line from Alice, the ring's fourth send from one request is refused | +| `d:seventh-refused` | The ring's exchange holds exactly 6 messages and the seventh is refused at the limit, while leg e's call is parked | +| `sweep:no-old-path` | None of [D2](DECISIONS.md#d2)'s `MESSAGE_NAMES_REMOVED` has a reference in what the CI sweep reads, each at its scope | + +**The rest of the plan** is [in DECISIONS](DECISIONS.md#the-plan-in-full): one browser journey, +the children's checks under a skip map, and the gap sweep. A failure becomes a child of FIX-2070 +blocking this issue, and the plan runs again as the closure rule says. + +## What changes + +```mermaid +sequenceDiagram + participant Al as Alice + participant HA as Helper A (files, calls) + participant R as Researcher + participant HB as Helper B's task + Al->>HA: ask Researcher the codename + HA->>R: call · row written + Note over HA,R: server killed before the dispatch, restarted + R-->>HA: lands once (c) · answer in the replayed request (a) + HA->>HB: files a task + HA->>R: second call, parked (Researcher waits on Alice) + HB-->>HA: midway question while the call waits (e) + HA->>HB: steer releases the hold + Al->>R: answers the stop · the call resolves once + Note over Al,R: the ring runs on its own line meanwhile (d) +``` + +Before, each arrow lived in its own child's check on a calm host; the closure runs them on one. + +**The sweep, one rule in the engine CI already runs** ([D2](DECISIONS.md#d2)): + +```diff + // goals/workforce-coordination/runs-on-one-task/rules.mjs ++ // FIX-2070's old message surface: no reference where the sweep reads, matched as ++ // PERSON_STOPS_REMOVED is, with the same `--control` plant ++ export const MESSAGE_NAMES_REMOVED = [ ++ "onWorkerMessage", "messageWorker", "MESSAGE_WORKER_TOOL", "WORKER_MESSAGE_ENTRY", ++ "workerMessageEntry", "sendWorkerMessage", "takesWorkerMessages", "workforceRelayTo", ++ "takes_tasks_only", "no_turn", "answerTo", ++ { name: "replyTo", files: /^packages\/(workforce|shift-manager)\/|^apps\/docs\/(docs|guides)\// }, // core keeps the word ++ ]; +``` + +**Docs:** no documentation impact. The closure ships checks; a page that doesn't hold when followed +is a finding against its child. + +**Stays as it is:** all product code; every child's check as it is on `main`, except that +`workers-message-each-other`'s own sweep reads D2's list instead of its copy. No legacy leg: +stores are reset between the refactor epics. + +## What's already here, and the debt + +| Already here | What this does with it | +|---|---| +| The epic's legs a to e and its sweep | **Builds on.** Walked together, not apart | +| `workers-message-each-other` (legs 1 to 8 on `main`, 9 and 10 to come) and `every-stop-is-one-inbox-item` j1 | **Composed.** Cast and world reused; five controls reused | +| `workers-message-each-other`'s `sweep` in `run.mts`, run only with that goal | **Replaced** by D2's rule, which it then reads | +| [FIX-2134](../FIX-2134/SPEC.md), the last closure, and `PERSON_STOPS_REMOVED` | **Kept** its shape; **extended** by one list beside it | +| Anything else grading the set | **None.** Searched `goals/` for `sendMessage`, `onMessageReceived`, `callId`, `closure` | + +| The debt | | +|---|---| +| **Removes** | One copy of the removed-names list | +| **Adds** | One goal and one CI rule. Permanent by design | +| **Leaves** | Kitchen-sink draws no `workforce:received` item; outside this epic, not graded | + +**Net: up, a little.** One goal and one list, for a promise that stays true. + +## Sign off + +**[The goal](#the-goal-and-how-well-know-its-met), at that size:** the five legs on one durable +host, colliding, on a real model and one commit. If wrong: the epic wraps on pieces that pass apart +and break together. + +**[The debt](#whats-already-here-and-the-debt), net up a little:** one goal and one CI rule. If +wrong: a CI rule that fails a later PR reusing one of twelve names. + +1. **[D1](DECISIONS.md#d1) · One walk where the legs collide, not the children's checks side by + side.** If wrong: a longer run whose failures are harder to split. +2. **[D2](DECISIONS.md#d2) · The removed names are one rule in the CI sweep, beside FIX-2134's.** + If wrong: one more list to edit when a name is reused. + +**Open: none of its own.** D1 is the one to weigh. Before the run: FIX-2240's leg 9 is not on +`main` yet ([Must be green first](DECISIONS.md#the-plan-in-full)). + +Improvement · closure (QA) · `goals/` and one CI rule · large, repeats per finding · epic +[FIX-2070](../../epics/FIX-2070/SPEC.md), closure · required · runs after every other child merges +(FIX-2238, FIX-2160, FIX-2240 done; FIX-2239, FIX-2241 in review) · Linear [FIX-2242](https://linear.app/fixpoint-labs/issue/FIX-2242) diff --git a/specs/issues/FIX-2242/figures/d1-legs-collide.svg b/specs/issues/FIX-2242/figures/d1-legs-collide.svg new file mode 100644 index 0000000000..a1cb465313 --- /dev/null +++ b/specs/issues/FIX-2242/figures/d1-legs-collide.svg @@ -0,0 +1,100 @@ + + + + + + + D1 · THE CALL + Does the closure walk the epic's legs together on one host, or re-run each child's leg alone? + + + + + CHOSEN + One walk, the legs collide + + INSTEAD OF + The children's checks side by side + + + + + + COLLISIONS + BETWEEN LEGS + ▸ DECIDES IT + + + They show + a call across a crash, a question to a parked filer + − + Can never happen + each leg runs on a calm host + + + + A FAILED RUN + − + Long, harder to split + a kill and a restart; every assertion tagged + + + Short, names one leg + and repeats what the children ran + + + + + EACH LEG ALONE + = + Proved + the children's checks, re-run in row 3 + = + Proved + the same + + + + + + LOCKS IN + A real model, a kill and a restart per attempt + and the whole run again for each finding + + + + + FLIPS IF + A host can't be killed mid-call in a goal run + then a and c walk on their own + + diff --git a/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg b/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg new file mode 100644 index 0000000000..d0a6bec855 --- /dev/null +++ b/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg @@ -0,0 +1,100 @@ + + + + + + + D2 · THE CALL + After the wrap, who notices if messageWorker or the relay comes back? + + + + + CHOSEN + One rule in the CI sweep + + INSTEAD OF + The sweep in one goal's run + + + + + + A LATER PR + BRINGS ONE BACK + ▸ DECIDES IT + + + That PR fails CI + a reviewer sees the list edited + − + Seen at the next closure + if anyone runs that goal again + + + + A NAME REUSED + − + The list is edited + replyTo scoped to Workforce, Shift Manager, pages + + + Costs nothing + between closures + + + + + ON THIS COMMIT + = + The same lines reported + one engine, with its plants + = + The same lines reported + the same + + + + + + LOCKS IN + Naming a removed name fails CI + until the PR edits the list, in view of a reviewer + + + + + FLIPS IF + The CI sweep stops reading source + then the goal's own sweep is all there is + + From 6ba3fffbb0e6a2c3fc824b12bed46fd99947da65 Mon Sep 17 00:00:00 2001 From: Claude Date: Wed, 28 Oct 2026 01:41:33 +0000 Subject: [PATCH 2/5] spec(FIX-2242): fold review: wait again graded in leg e, j1 reads a call as a tool result, rows 3 and 4 trimmed Co-Authored-By: Claude Opus 5.5 Claude-Session: https://claude.ai/code/session_01NczDiEtVsnByx4XbajAf5k --- specs/issues/FIX-2242/DECISIONS.md | 11 ++++++++--- specs/issues/FIX-2242/SPEC.md | 11 +++++++---- 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/specs/issues/FIX-2242/DECISIONS.md b/specs/issues/FIX-2242/DECISIONS.md index b3596568fe..4026f05ad3 100644 --- a/specs/issues/FIX-2242/DECISIONS.md +++ b/specs/issues/FIX-2242/DECISIONS.md @@ -56,9 +56,9 @@ the PR adding the rule updates that fixture to a refusal that exists. | # | What runs | PASS means | |---|---|---| | 1 | Legs a to e and the sweep ([the signal](SPEC.md#the-signal)) | Each green; each control red where named | -| 2 | **j1** *uses Shift Manager*, in a browser as Alice after the walk: Helper A's conversation shows Researcher's answer once, as its own item, with its sender and the message it answers (`j1:own-item`); the Messages view lists the crashed call once, `delivered`, the parked call `answered`, and the ring's seventh `refused` (`j1:states`); each side's link opens its session (`j1:opens-side`). The other four teams' *after* is walked by leg a (builds talkers), e (files work), c and d (production) and the sweep (maintains); each mapping is printed | Each green | -| 3 | Every child's goal check with its controls, read from Linear at run time and printed: `workers-message-each-other` legs 1 to 10c with every control, FIX-2240's leg 9 included; `every-stop-is-one-inbox-item` j1 and j1b with FIX-2241's controls, and its other legs, since FIX-2241 changed its Lab. A check row 1 walks in full is skipped, printed beside the leg that covers it; one walked in part runs. #4021's two budget cases are CI tests | Each PASS beside its documented FAIL | -| 4 | The gap sweep: each [coordination seam](../../epics/FIX-2070/DECISIONS.md#coordination-seams) from both sides (on the send path, a `{ task }` target resolved before the limits, the call's wait after the dispatch, the resend through the wake; every field on the message row written by its owner and read by its reader; every worker flow, coordinators and the coder included, carrying `onMessageReceived`; the question sent through `sendMessage`'s server path, outside the budget; the Messages view's states, `undeliverable` and a waiting call included; FIX-2071's names untouched); [`messaging-workers.md`](../../../apps/docs/docs/workforce/messaging-workers.md) and the Workforce README followed as written, the page's call example run as a goal-local worker; each epic *not done if* state absent; the *Not doing* line on a session waiting on a user checked as a queue | Every seam two-sided | +| 2 | **j1** *uses Shift Manager*, in a browser as Alice after the walk: Researcher's conversation shows Helper A's call once, as its own item, with its sender and its call mark, and Helper A's conversation shows the call's tool line answered with the reply, never a second received item (`j1:own-item`); the Messages view lists the crashed call once, `delivered`, the parked call `answered`, and the ring's seventh `refused` (`j1:states`); each side's link opens its session (`j1:opens-side`). j1 grades only the screen: what it reads, row 1 has already graded in the store, so a j1 failure with row 1 green is a drawing bug. The other four teams' *after* is walked by leg a (builds talkers), e (files work), c and d (production) and the sweep (maintains); each mapping is printed | Each green | +| 3 | Every child's goal check with its controls, read from Linear at run time and printed: `workers-message-each-other` legs 1 to 10c with every control, FIX-2240's leg 9 included; `every-stop-is-one-inbox-item` j1 and j1b with FIX-2241's controls, and its other legs, since FIX-2241 changed its Lab. **None is skipped:** row 1 walks the legs together in its own scenes, never a child's legs as that child wrote them. #4021's two budget cases are CI tests | Each PASS beside its documented FAIL | +| 4 | The gap sweep: each [coordination seam](../../epics/FIX-2070/DECISIONS.md#coordination-seams) from both sides, as its row names them; [`messaging-workers.md`](../../../apps/docs/docs/workforce/messaging-workers.md) and the Workforce README followed as written; each epic *not done if* state absent | Every seam two-sided | **Must be green first:** every other child merged with its goal's PASS and its controls' FAIL; #4021 merged; the D2 rule in CI. **FIX-2240's leg 9 is not on `main`:** #4008 merged with its @@ -100,7 +100,12 @@ blocking this issue, since leg c reuses its faults and `no-debt`. filer parked on a call; FIX-2134's shape and FIX-2087's D1 kept; the sweep added to the CI engine beside `PERSON_STOPS_REMOVED`. Found on `main`: FIX-2240's leg 9 never landed, and a shift-manager test still names `takes_tasks_only`. +- **Review** — leg e waits again by call id (`e:kept-once`), j1 reads a call as a tool result, rows 3 and 4 trimmed (Codex, Cursor). ## Notes from review +- **Codex, folded:** `j1:own-item` looked for the call's answer as a received item, which `a:no-relay` forbids; it now reads the call in Researcher's conversation and the answered tool line in Helper A's. `send-again` graded `a:one-place`, which never waits again; leg e now waits again by call id and `e:kept-once` carries the control. +- **Cursor, folded:** row 3 says nothing is skipped and why; row 4 points at the seam table instead of restating it; j1 grades the screen only; the docs line says the pages are followed, not written. +- **Cursor, kept:** the sequence diagram is the required *What changes* figure, and the flowchart is the goal's; the figures' CSS is the trade-off template's, copied as every spec does. + **Open: none of its own.** diff --git a/specs/issues/FIX-2242/SPEC.md b/specs/issues/FIX-2242/SPEC.md index 6e1a3eb555..b1c94ab868 100644 --- a/specs/issues/FIX-2242/SPEC.md +++ b/specs/issues/FIX-2242/SPEC.md @@ -45,7 +45,7 @@ The walk reads sessions and the log through public actions; each control breaks | **Signal** | Every `:` in [the signal table](#the-signal) passes | | **Input** | DevTeam served by Shift Manager's command over a SQLite store file (`DEVTEAM_STORE`), killed and restarted on the same file; codenames and the brief's fact picked at run time | | **Anti-game** | Assert on sessions' stored items, the message log and `getTask`, read as Alice; nothing seeded; sent messages counted from each turn's own tool outputs; a fault that never fired is a setup failure, not a grade | -| **Control that must fail** | Today's `main` before FIX-2238 (`e2c190e`): legs a, b, c and e, and the sweep · `no-debt` (FIX-2240's) **`c:lands-once`** · `send-again` (FIX-2239's) **`a:one-place`** · `task-to-conversation` (FIX-2160's) **`b:in-task-session`** · `no-question-message` (FIX-2241's) **`e:filer-hears`** · `no-hop-cap` (FIX-2238's) **`d:seventh-refused`** · `call-holds-session`, new: a parked call keeps its session's place **`e:heard-while-calling`** · the sweep's `--control` its plants | +| **Control that must fail** | Today's `main` before FIX-2238 (`e2c190e`): legs a, b, c and e, and the sweep · `no-debt` (FIX-2240's) **`c:lands-once`** · `send-again` (FIX-2239's) **`e:kept-once`** · `task-to-conversation` (FIX-2160's) **`b:in-task-session`** · `no-question-message` (FIX-2241's) **`e:filer-hears`** · `no-hop-cap` (FIX-2238's) **`d:seventh-refused`** · `call-holds-session`, new: a parked call keeps its session's place **`e:heard-while-calling`** · the sweep's `--control` its plants | ### The signal @@ -57,7 +57,8 @@ The walk reads sessions and the log through public actions; each control breaks | `a:no-relay` | Researcher's turn sends no message back, and Helper A's conversation gains no received item from it | | `e:filer-hears` | Helper A files Helper B a task whose brief needs a fact only Helper A holds; Helper B asks it midway; the question arrives at Helper A's `onMessageReceived` in the session that filed the task | | `e:heard-while-calling` | It arrives while Helper A's second call, into Researcher's conversation holding a stop for Alice, is parked: the received item is stored before the call resolves | -| `e:steer-releases` | Helper A's steer releases the hold; the task finishes on attempt 1; the call then resolves once, answered after Alice's answer, or kept and returned when Helper A waits again by call id | +| `e:steer-releases` | Helper A's steer releases the hold, and the task finishes on attempt 1 | +| `e:kept-once` | The second call's deadline passes before Alice answers Researcher's stop, so it returns `running`; after her answer, Helper A waits again by call id and gets Researcher's reply. The log holds one row and one budget place for that call | | `b:in-task-session` | A message to Helper B's task while its turn runs lands in that task's session after the turn ends; one sent after the task finished is answered there from the task's history | | `b:task-unchanged` | The row's status, output, attempts and session are the same before and after both | | `d:per-request` | Started on its own line from Alice, the ring's fourth send from one request is refused | @@ -84,7 +85,9 @@ sequenceDiagram HA->>R: second call, parked (Researcher waits on Alice) HB-->>HA: midway question while the call waits (e) HA->>HB: steer releases the hold - Al->>R: answers the stop · the call resolves once + R-->>HA: deadline passes · running + Al->>R: answers the stop + HA->>R: waits again by call id · the reply, once Note over Al,R: the ring runs on its own line meanwhile (d) ``` @@ -104,7 +107,7 @@ Before, each arrow lived in its own child's check on a calm host; the closure ru + ]; ``` -**Docs:** no documentation impact. The closure ships checks; a page that doesn't hold when followed +**Docs:** no documentation impact: no page is written or changed. Row 4 follows the pages FIX-2238 and FIX-2239 published, and the sweep reads them; a page that doesn't hold when followed is a finding against its child. **Stays as it is:** all product code; every child's check as it is on `main`, except that From f226cfd972325bc0e0889ed43c21a1c5cc125ef3 Mon Sep 17 00:00:00 2001 From: Claude Date: Wed, 28 Oct 2026 01:49:14 +0000 Subject: [PATCH 3/5] spec(FIX-2242): D2 lives in the vocabulary sweep's RETIRED, with the full name list; leg 9's owner and the new control's file named; the epic's leg a reworded Co-Authored-By: Claude Opus 5.5 Claude-Session: https://claude.ai/code/session_01NczDiEtVsnByx4XbajAf5k --- specs/epics/FIX-2070/SPEC.md | 4 +- specs/issues/FIX-2242/DECISIONS.md | 26 +++++++------ specs/issues/FIX-2242/SPEC.md | 38 +++++++++++-------- .../FIX-2242/figures/d2-sweep-in-ci.svg | 4 +- 4 files changed, 41 insertions(+), 31 deletions(-) diff --git a/specs/epics/FIX-2070/SPEC.md b/specs/epics/FIX-2070/SPEC.md index ede4024688..7a93723b86 100644 --- a/specs/epics/FIX-2070/SPEC.md +++ b/specs/epics/FIX-2070/SPEC.md @@ -38,7 +38,7 @@ plus the message half of the *2 · One Inbox* question row, which FIX-2058 moved ```mermaid flowchart LR - A["Shift Manager · real model · durable file-backed host"] --> L1["leg a · writer calls coder"] + A["Shift Manager · real model · durable file-backed host"] --> L1["leg a · a worker calls another"] A --> L2["leg b · a message to a task's session, open and finished"] A --> L3["leg c · restart between the log write and the delivery"] A --> L4["leg d · a ring of workers passing messages on"] @@ -60,7 +60,7 @@ tries), and a filer hears no question. | How we verify | | |---|---| | **Goal check** | The closure issue's (FIX-2242), real model, in Shift Manager, on one `main` commit after every other child merges | -| **Signal** | Leg a: writer's request reads coder's answer as the call's result; past a short deadline it resumes without it, and waiting again returns the kept result. Leg b: the message runs as `onMessageReceived` in the task's session, with its history; an open task's turn finishes first, and the task's status, output and attempts don't change. Leg c: the server stops after the log row is written and before the dispatch; after the restart the recipient sees the message once. Leg d: the fourth message from one request and the seventh from one typed message are refused. Leg e: the midway question arrives at the filer's `onMessageReceived` in the session that filed the task; its steer releases the hold. A sweep: the removed names have no references in package source | +| **Signal** | Leg a: the caller's request reads the other worker's answer as the call's result (the coder hears only in its tasks' sessions, FIX-2238 D1, so it is reached by `{ task }`); past a short deadline it resumes without it, and waiting again returns the kept result. Leg b: the message runs as `onMessageReceived` in the task's session, with its history; an open task's turn finishes first, and the task's status, output and attempts don't change. Leg c: the server stops after the log row is written and before the dispatch; after the restart the recipient sees the message once. Leg d: the fourth message from one request and the seventh from one typed message are refused. Leg e: the midway question arrives at the filer's `onMessageReceived` in the session that filed the task; its steer releases the hold. A sweep: the removed names have no references in package source | | **Anti-game** | Assert on what the recipient's session and the message log show, read through public actions; never on a child's own helper | | **Control that must fail** | Today's `main`: legs a, b, c and e, and the sweep. Every other check names its control, and each is run red at least once | diff --git a/specs/issues/FIX-2242/DECISIONS.md b/specs/issues/FIX-2242/DECISIONS.md index 4026f05ad3..b1cae5cb44 100644 --- a/specs/issues/FIX-2242/DECISIONS.md +++ b/specs/issues/FIX-2242/DECISIONS.md @@ -12,7 +12,7 @@ Two calls are this spec's: how the walk puts the legs together, and where the sw flowchart TD I["FIX-2242"] --> D1["D1 · one walk where the legs collide, one durable host"] D1 -.->|"rejected · each child already walks its leg alone"| X1["the children's checks re-run side by side"] - I --> D2["D2 · the removed names are one rule in the CI sweep"] + I --> D2["D2 · the removed names are rows in the vocabulary sweep"] D2 -.->|"rejected · a later PR brings a name back unseen"| X2["the sweep in one goal's run only"] ``` @@ -35,21 +35,22 @@ It comes down to collisions: a check that walks one leg on a calm host can never Then a and c walk on their own, and e meets a parked call that never crosses a restart. -## D2 · The removed names are one rule in the CI sweep, beside FIX-2134's, not a check one goal runs +## D2 · The removed names are two rows in the vocabulary sweep's `RETIRED`, not a check one goal runs | | | |---|---| | **Instead of** | `workers-message-each-other`'s own sweep (`run.mts`, `REMOVED`), which runs only when that goal runs, over `packages/*/src` | -| **Because** | The epic promises one entry and one tool. The likeliest way to break that after the wrap is a later PR bringing back `messageWorker`, the relay or a refusal under its old name, and only CI sees every PR. FIX-2087's engine already runs in CI with plants, and FIX-2134 added `PERSON_STOPS_REMOVED` beside its rows, so this is one more list in the same pass, not a second sweep. The goal's sweep then reads this list, so there is one copy | -| **Locks in** | A PR naming one of the twelve where the sweep reads (package source and tests, the pages a user reads; comments skipped) fails CI until it edits the list. `replyTo` is a word `core`'s dispatcher and tests still use for their own fields, so it is matched only in `packages/workforce`, `packages/shift-manager` and the pages, as FIX-2134 scoped `answerQuestion` | +| **Because** | The epic promises one entry and one tool. The likeliest way to break that after the wrap is a later PR bringing back `messageWorker`, the relay or a refusal under its old name, and only CI sees every PR. FIX-2071's renames table (#3985, ER-3) already names `one-word-per-thing`'s `RETIRED` as these names' home, so putting them in FIX-2087's engine too would make two lists with two `replyTo` scopes. `RETIRED` runs in CI with a `--control` plant, and the goal's sweep then reads its rows, so there is one copy. #3985 moves the row's owner from FIX-2249 to this issue | +| **Locks in** | A PR naming one of the nineteen in package code (tests included) or a published page fails CI until it edits the rows. `replyTo` is a word `core`'s dispatcher and tests still use for their own fields, so its row reads only `packages/workforce`, `packages/shift-manager` and the pages. The entry's first name is swept as `receiveProblem`, `receiveEntry` and `RECEIVE_ENTRY`, never a bare `receive` | -![D2: one rule in the CI sweep, chosen, beside the sweep in one goal's run; decided by a later PR bringing a removed name back](figures/d2-sweep-in-ci.svg) +![D2: rows in the vocabulary sweep, chosen, beside the sweep in one goal's run; decided by a later PR bringing a removed name back](figures/d2-sweep-in-ci.svg) It comes down to a later PR: swept only by one goal, a returned `messageWorker` is seen at the next closure, if ever. **On `main` today the list is not yet green:** `packages/shift-manager/test/messages-view.test.tsx` -still renders a `takes_tasks_only` refusal, which FIX-2238 removed. The CI sweep reads tests, so -the PR adding the rule updates that fixture to a refusal that exists. +still renders a `takes_tasks_only` refusal (line 126), which FIX-2238 removed, and #4020, #4021 +and #4022 carry it too. `RETIRED` reads tests, so the PR adding the rows updates that fixture to a +refusal that exists, on the commit it lands on. ## The plan in full @@ -63,8 +64,8 @@ the PR adding the rule updates that fixture to a refusal that exists. **Must be green first:** every other child merged with its goal's PASS and its controls' FAIL; #4021 merged; the D2 rule in CI. **FIX-2240's leg 9 is not on `main`:** #4008 merged with its model-free tests and without the goal leg its spec names (`9:crash-after-row` and the rest, with -`no-debt` and the other four controls). It lands before the run, as its own child of FIX-2070 -blocking this issue, since leg c reuses its faults and `no-debt`. +`no-debt` and the other four controls). FIX-2240's approved spec still owes it, so it is filed as +a child of FIX-2070 ("FIX-2240's goal leg 9 on the real model") and lands before the run, blocking this issue, since leg c reuses its faults and `no-debt`. ## Decided, not asked @@ -78,7 +79,9 @@ blocking this issue, since leg c reuses its faults and `no-debt`. - **[FIX-2087 D1](../FIX-2087/DECISIONS.md#d1), inherited:** a structural miss on any attempt is a finding; only a judgment miss (what the model chose to write) earns another attempt. - **One new control, `call-holds-session`:** a parked call keeps its session's place, so a - question to the filer waits for the call to end. It is the one collision no child's control + question to the filer waits for the call to end. It patches FIX-2239's call park + (`packages/workforce/src/messages/call.ts` in #4020) to hold the turn in-process instead of + parking on the wait gate. It is the one collision no child's control breaks. A patch that never fired fails the build. - **Today's `main` is `e2c190e`,** the commit before FIX-2238. Leg d can pass there, since FIX-1989's hop limit already stopped a ring at 6; the epic names a, b, c and e. @@ -100,12 +103,13 @@ blocking this issue, since leg c reuses its faults and `no-debt`. filer parked on a call; FIX-2134's shape and FIX-2087's D1 kept; the sweep added to the CI engine beside `PERSON_STOPS_REMOVED`. Found on `main`: FIX-2240's leg 9 never landed, and a shift-manager test still names `takes_tasks_only`. -- **Review** — leg e waits again by call id (`e:kept-once`), j1 reads a call as a tool result, rows 3 and 4 trimmed (Codex, Cursor). +- **Review** — leg e waits again by call id (`e:kept-once`), j1 reads a call as a tool result, rows 3 and 4 trimmed (Codex, Cursor); D2 moved into the vocabulary sweep's `RETIRED`, and the epic's leg a reworded (FSD Architect). ## Notes from review - **Codex, folded:** `j1:own-item` looked for the call's answer as a received item, which `a:no-relay` forbids; it now reads the call in Researcher's conversation and the answered tool line in Helper A's. `send-again` graded `a:one-place`, which never waits again; leg e now waits again by call id and `e:kept-once` carries the control. - **Cursor, folded:** row 3 says nothing is skipped and why; row 4 points at the seam table instead of restating it; j1 grades the screen only; the docs line says the pages are followed, not written. +- **FSD Architect, folded:** D2 moved from FIX-2087's engine to `one-word-per-thing`'s `RETIRED`, the home #3985's renames table already gives these names, so there is one list and one `replyTo` scope; the list gained `SendWorkerMessage`, the two input schemas, `WORKER_MESSAGES_SEEN_STATE` and the entry's first names (`receiveProblem`, `receiveEntry`, `RECEIVE_ENTRY`); the fixture note corrected (FIX-2087's engine skips tests, `RETIRED` reads them); leg 9's owner and `call-holds-session`'s file named; the epic's "writer calls coder" amended. - **Cursor, kept:** the sequence diagram is the required *What changes* figure, and the flowchart is the goal's; the figures' CSS is the trade-off template's, copied as every spec does. **Open: none of its own.** diff --git a/specs/issues/FIX-2242/SPEC.md b/specs/issues/FIX-2242/SPEC.md index b1c94ab868..fa5cf89d58 100644 --- a/specs/issues/FIX-2242/SPEC.md +++ b/specs/issues/FIX-2242/SPEC.md @@ -63,7 +63,7 @@ The walk reads sessions and the log through public actions; each control breaks | `b:task-unchanged` | The row's status, output, attempts and session are the same before and after both | | `d:per-request` | Started on its own line from Alice, the ring's fourth send from one request is refused | | `d:seventh-refused` | The ring's exchange holds exactly 6 messages and the seventh is refused at the limit, while leg e's call is parked | -| `sweep:no-old-path` | None of [D2](DECISIONS.md#d2)'s `MESSAGE_NAMES_REMOVED` has a reference in what the CI sweep reads, each at its scope | +| `sweep:no-old-path` | None of [D2](DECISIONS.md#d2)'s `message-names` and `message-reply-to` rows has a hit in what the vocabulary sweep reads, tests included, each at its scope | **The rest of the plan** is [in DECISIONS](DECISIONS.md#the-plan-in-full): one browser journey, the children's checks under a skip map, and the gap sweep. A failure becomes a child of FIX-2070 @@ -93,25 +93,30 @@ sequenceDiagram Before, each arrow lived in its own child's check on a calm host; the closure runs them on one. -**The sweep, one rule in the engine CI already runs** ([D2](DECISIONS.md#d2)): +**The sweep, two rows in the vocabulary sweep CI already runs** ([D2](DECISIONS.md#d2)), the home +FIX-2071's renames table (#3985) gives these names: ```diff - // goals/workforce-coordination/runs-on-one-task/rules.mjs -+ // FIX-2070's old message surface: no reference where the sweep reads, matched as -+ // PERSON_STOPS_REMOVED is, with the same `--control` plant -+ export const MESSAGE_NAMES_REMOVED = [ -+ "onWorkerMessage", "messageWorker", "MESSAGE_WORKER_TOOL", "WORKER_MESSAGE_ENTRY", -+ "workerMessageEntry", "sendWorkerMessage", "takesWorkerMessages", "workforceRelayTo", -+ "takes_tasks_only", "no_turn", "answerTo", -+ { name: "replyTo", files: /^packages\/(workforce|shift-manager)\/|^apps\/docs\/(docs|guides)\// }, // core keeps the word -+ ]; + // goals/vocabulary/one-word-per-thing/rules.mjs + export const RETIRED = [ + … ++ { id: "message-names", files: CODE_OR_PUBLISHED, // FIX-2070's retired message surface ++ pattern: /\b(onWorkerMessage|messageWorker|MESSAGE_WORKER_TOOL|WORKER_MESSAGE_ENTRY|workerMessageEntry| ++ sendWorkerMessage|SendWorkerMessage|takesWorkerMessages|messageWorkerInputSchema| ++ workerMessageInputSchema|WORKER_MESSAGES_SEEN_STATE|workforceRelayTo|takes_tasks_only|no_turn| ++ answerTo|receiveProblem|receiveEntry|RECEIVE_ENTRY)\b/ }, // never a bare `receive` ++ { id: "message-reply-to", // core's dispatcher keeps the word ++ files: new RegExp(`^packages\\/(workforce|shift-manager)\\/.*\\.(ts|tsx|mts)$|${PUBLISHED.source}`), ++ pattern: /\breplyTo\b/ }, + ]; ``` **Docs:** no documentation impact: no page is written or changed. Row 4 follows the pages FIX-2238 and FIX-2239 published, and the sweep reads them; a page that doesn't hold when followed is a finding against its child. **Stays as it is:** all product code; every child's check as it is on `main`, except that -`workers-message-each-other`'s own sweep reads D2's list instead of its copy. No legacy leg: +`workers-message-each-other`'s own sweep reads D2's rows instead of its copy, and one Shift +Manager test fixture that still renders a removed refusal. No legacy leg: stores are reset between the refactor epics. ## What's already here, and the debt @@ -121,7 +126,8 @@ stores are reset between the refactor epics. | The epic's legs a to e and its sweep | **Builds on.** Walked together, not apart | | `workers-message-each-other` (legs 1 to 8 on `main`, 9 and 10 to come) and `every-stop-is-one-inbox-item` j1 | **Composed.** Cast and world reused; five controls reused | | `workers-message-each-other`'s `sweep` in `run.mts`, run only with that goal | **Replaced** by D2's rule, which it then reads | -| [FIX-2134](../FIX-2134/SPEC.md), the last closure, and `PERSON_STOPS_REMOVED` | **Kept** its shape; **extended** by one list beside it | +| [FIX-2134](../FIX-2134/SPEC.md), the last closure | **Kept** its shape | +| The vocabulary sweep's `RETIRED`, where FIX-2071's renames table (#3985, ER-3) puts these names | **Extended** by two rows; #3985's row moves from FIX-2249 to this issue | | Anything else grading the set | **None.** Searched `goals/` for `sendMessage`, `onMessageReceived`, `callId`, `closure` | | The debt | | @@ -139,12 +145,12 @@ host, colliding, on a real model and one commit. If wrong: the epic wraps on pie and break together. **[The debt](#whats-already-here-and-the-debt), net up a little:** one goal and one CI rule. If -wrong: a CI rule that fails a later PR reusing one of twelve names. +wrong: a CI rule that fails a later PR reusing one of nineteen names. 1. **[D1](DECISIONS.md#d1) · One walk where the legs collide, not the children's checks side by side.** If wrong: a longer run whose failures are harder to split. -2. **[D2](DECISIONS.md#d2) · The removed names are one rule in the CI sweep, beside FIX-2134's.** - If wrong: one more list to edit when a name is reused. +2. **[D2](DECISIONS.md#d2) · The removed names are two rows in the vocabulary sweep's `RETIRED`,** + the home FIX-2071's renames table already gives them. If wrong: one more row to edit when a name is reused. **Open: none of its own.** D1 is the one to weigh. Before the run: FIX-2240's leg 9 is not on `main` yet ([Must be green first](DECISIONS.md#the-plan-in-full)). diff --git a/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg b/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg index d0a6bec855..3da2bf7d40 100644 --- a/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg +++ b/specs/issues/FIX-2242/figures/d2-sweep-in-ci.svg @@ -1,4 +1,4 @@ -