Skip to content

project(orchestration-primitives): FIX-2071 wraps — the renames landed; four calls decided once - #4254

Merged
jhoffner merged 3 commits into
project/orchestration-primitivesfrom
claude/project-thread-q4yqvs
Oct 31, 2026
Merged

jhoffner merged 3 commits into
project/orchestration-primitivesfrom
claude/project-thread-q4yqvs

Conversation

@jhoffner

Copy link
Copy Markdown
Collaborator

Requested by Jake · project thread

The wrap update for FIX-2071 (coordination renames) on the Orchestration Primitives project-spec. It targets project/orchestration-primitives, not main. Merging it lands the update on the never-merged project PR #3102.

What moved

  • Epics table, re-derived from Linear on Oct 30. FIX-2071 is done (wrapped Oct 30). FIX-2070 is in flight: its builds landed and closure FIX-2242 is running. FIX-2092 is in flight with its epic spec spec(epic): phased workflows, a feature a person can follow phase by phase (FIX-2092) #3970 in review. The count is now 9 done · 2 in flight · 0 filed, not started. The mermaid graph marks FIX-2071 done.
  • Arc (figures/arc.svg). FIX-2071's bar closes (Oct 21 to Oct 30). FIX-2070 and FIX-2092 get in-flight bars. The now line moves to Oct 30. Rendered and checked in light and dark.
  • Plan. A new "What FIX-2071 left open" section: the lessons draft docs(lessons): FIX-2071 epic wrap, cycle 28 ledger #4245 (two asks for the owner) and fold candidates F1–F7. The "waits on FIX-2070" lines are replaced.
  • DECISIONS.md → decided once, four rows from FIX-2071:
    • D1: a renamed name is refused loudly, never kept as an alias.
    • ER-2: every rename ships a committed codemod in scripts/codemod/.
    • ER-3: the RETIRED rows in goals/vocabulary/one-word-per-thing only ever grow.
    • D2: specs/epics/FIX-2071/renames.md is the vocabulary contract.

Outcome, PD-1 to PD-3 and the rules are unchanged.

Evidence (checked on main at 863a0e1)

The Linear project's content is re-mirrored from this head.


Generated by Claude Code

…d; four calls decided once

FIX-2071's row goes to done (seven children and closure #4244 merged,
Linear Done Oct 30) and its arc bar closes. Status re-derived from
Linear: FIX-2070 and FIX-2092 are in flight. The now line moves to
Oct 30.

Decided once, from FIX-2071: D1 (a renamed name is refused, never an
alias), ER-2 (every rename ships a committed codemod), ER-3 (RETIRED
rows only grow), and D2's rename table as the vocabulary contract.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NczDiDaAHdCicsHDpBceid
@jhoffner jhoffner self-assigned this Oct 30, 2026
@vercel

vercel Bot commented Oct 30, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

3 Skipped Deployments
Project Deployment Actions Updated
flow-state-dev-docs Ignored Ignored Oct 30, 2026 8:33pm UTC
flow-state-dev-kitchen-sink Ignored Ignored Oct 30, 2026 8:33pm UTC
flow-state-dev-ui Ignored Ignored Preview Oct 30, 2026 8:33pm UTC

Request Review

@linear-code

linear-code Bot commented Oct 30, 2026

Copy link
Copy Markdown

FIX-2071

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 30, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-30T20:34:29.290291Z 747a152 New commits
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

Copy link
Copy Markdown
Collaborator Author

Second look: PR #4254

Verdict: approach is right, ready after two small stale-text fixes. No refactor warranted.

This is a derived-status refresh of a project spec (docs/figures only, no code), so the leaner path is already taken. It re-derives from Linear and adds four decision rows, and it leaves the outcome, PD-1 to PD-3 and the rules alone.

Checked and fine

  • Counts: 9 done · 2 in flight · 0 filed is 11 epics. That matches the table, the mermaid done class (D2071 added) and the "Now" row.
  • arc.svg: the geometry holds against the stated formula x = 180 + 720/199·days. Oct 21 is day 189, which gives 863.8. Oct 30 is day 198, which gives 896.4. Bar width 32.6 reaches the now line. The aria-label matches the lanes.
  • The new D1 row's "no-deprecation-release row above" reference resolves (DECISIONS.md line 85).
  • Each new decision row carries an owner/source and an anchor link.

Fix before merge (stale text this PR makes wrong)

  1. PLAN.md:27 still describes FIX-2092 as "A design to prove by POC before its spec". The epic spec is now in review (spec(epic): phased workflows, a feature a person can follow phase by phase (FIX-2092) #3970), and the table you edited says so. Use one status in both places.
  2. PLAN.md:28 says FIX-2071 is "FIX-2056's read for retired rows removed no earlier than here". That is future-tense dependency language for an epic now done. Say whether the read was removed. If it wasn't, it belongs in the "What FIX-2071 left open" table.

Nits

Architecture note (not for this PR)
The epic counts are hand-copied into SPEC "Now", the SPEC paragraph under the table, and PLAN prose. Each wrap updates all three by hand, and this PR touches all three. That is a drift risk, since a skipped edit leaves them disagreeing. Consider keeping one derived line (the table) and having the other two refer to it. Don't fold that into a wrap PR.


Generated by Claude Code

Copy link
Copy Markdown
Collaborator Author

Process guards is red because of where this PR comes from, not what it changes. scripts/validate-spec-folder.mjs refuses any spec/_projects content unless the PR's head branch is project/*. This PR's head is claude/project-thread-q4yqvs, with base project/orchestration-primitives. The guard has no exemption for a PR into a project branch, so it fails on spec/_projects even though the base already holds that folder.

There is no fix inside this PR. Moving the files would defeat the PR. Changing the guard is out of scope for a status update. The same commit (c28c289) passes the guard if it is pushed straight to project/orchestration-primitives, which is the project-agent's normal route. Merging this PR as it stands lands the same content. A re-run would only repeat the result, so I haven't re-run it.


Generated by Claude Code

@cursor cursor Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

/simplify review

Verdict: Approve. This is a spec-only project refresh after FIX-2071 wrapped, and it follows the same shape as the FIX-2058 wrap (#684fbbc): mirror binding decisions into DECISIONS.md, add a What FIX-2071 left open section on PLAN.md, bump the live epics table and kill line on SPEC.md, and hand-update figures/arc.svg. That four-surface sync is deliberate for this project-spec set, not accidental scope creep in this PR.

Did the approach go too far? The underlying FIX-2071 epic was heavy (codemods, vocabulary, refusal semantics). This PR mostly records outcomes. The one place I’d consider a lighter alternative (optional, not blocking) is how much FIX-2071 process lands in the project decided once table versus staying in specs/epics/FIX-2071/ with a link — see inline on codemod / RETIRED rows.

Cleaner alternatives considered (for future refreshes, not a rewrite of this PR):

Idea Tradeoff
Single canonical status — treat SPEC.md → The epics as source of truth; shorten arc aria-label and keep PLAN on ordering/hand-offs only Less duplicate prose on every “now line” bump; arc becomes more schematic
Generate arc.svg from a small data file (dates + state per lane) Up-front script vs repeated magic coordinates (896.4, bar widths); no third-party package required, but repo today is hand-maintained SVG across project figures
Trim the seven-child PR laundry list in PLAN to match FIX-2056/2058 “left open” sections (closure + carried-forward table only) History stays on epic spec / Linear; project PLAN stays scannable

Skipped as out of scope / uncertain: Regenerating mermaid styling for in-flight epics (edges still read D2070 → D2071 as strict sequencing while prose correctly says 2071 wrapped before 2070’s closure). Worth a follow-up pass if the diagram is meant to track partial deps, not just wrap order. Terminology in The read still says “hold” post-rename — likely a separate terminology sync, not required for this status refresh.

Performance: Static SVG, no SMIL; ~828-char aria-label is maintainability/a11y verbosity, not runtime cost.

No code changes requested from this automation run.

Open in Web View Automation 

Sent by Cursor Automation: Review For Best Approach

…X-2071 consumes FIX-2070's builds

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NczDiDaAHdCicsHDpBceid

@cursor cursor Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Snob. The wrap note is longer than the contract it carries.

PLAN.md lists seven merged child PRs, then a two-row table whose rows both point at lessons draft #4245, then a sentence that the four decided-once rows live in DECISIONS.md. The epics table already links the closure (#4244) and that draft. The order blurb also restates that FIX-2070 and FIX-2092 are in flight.

Cut those. Keep the closure citation, the two owner asks, and F1–F7. Leave the four decision rows, the arc, and the epics table alone.

Shown on draft #4256 (poc/shorter-fix-2071-left-open, c7fff8495). Spec-figure checks still pass. PLAN 1260 → 1216 words.

That draft does not merge. Agent on this PR: fold the subtraction, then close the POC unmerged after this one lands.

Open in Web View Automation 

Sent by Cursor Automation: Code Snob

…e closure and lessons once

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NczDiDaAHdCicsHDpBceid

Copy link
Copy Markdown
Collaborator Author

Pushed 747a152, which folds both Cursor reviews. Process guards will stay red for the branch-name reason above until the owner decides the route. The Linear mirror now matches 747a152.


Generated by Claude Code

@jhoffner
jhoffner merged commit 7927905 into project/orchestration-primitives Oct 31, 2026
15 of 16 checks passed
@jhoffner
jhoffner deleted the claude/project-thread-q4yqvs branch October 31, 2026 00:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants