IW-607: A resumed phase-pr judges the pushed head, not a stale one #692

Merged

mph merged 22 commits from IW-607 into main 2026-10-07 07:31:11 +00:00

Conversation 2 Commits 22 Files changed 15 +1092 -10

claude commented 2026-10-03 01:19:16 +00:00

Member

Copy link

Summary

A resumed phase-pr pushes the branch onto a pull request that is already open and then gates CI at once. When the forge was slow to move the PR head to the pushed commit, the gate judged the old head: a stale red used up a recovery attempt, and a stale green handed over (or merged with --batch) a commit CI never judged.

Defects fixed

• Phase 1 (#689): phase-pr reads the local HEAD after its push and passes it to the gate as CiGateTarget.expectedHead. The gate now waits while the forge reports another head (the existing IW-601 guard), and judges only the pushed commit. A HEAD that cannot be read leaves the gate as before (analysis decision D1).

Out of scope

• kanon awaitCi has the same race. It needs a ci-gate option for the expected head — filed as #690.
• A test gap on the red-report harvest path — filed as #691.

Testing

• Regression tests in core/test/PhasePrHarnessTest.scala: stale red, stale green, a resume that pushed nothing new, a timeout while the forge keeps the old head, a first push, and an unreadable HEAD.
• CI green on the phase PR (#689).

Release notes

See project-management/issues/IW-607/release-notes.md (Czech).

🤖 Generated with Claude Code

Closes #607

## Summary A resumed `phase-pr` pushes the branch onto a pull request that is already open and then gates CI at once. When the forge was slow to move the PR head to the pushed commit, the gate judged the old head: a stale red used up a recovery attempt, and a stale green handed over (or merged with `--batch`) a commit CI never judged. ## Defects fixed - **Phase 1 (#689):** `phase-pr` reads the local HEAD after its push and passes it to the gate as `CiGateTarget.expectedHead`. The gate now waits while the forge reports another head (the existing IW-601 guard), and judges only the pushed commit. A HEAD that cannot be read leaves the gate as before (analysis decision D1). ## Out of scope - kanon `awaitCi` has the same race. It needs a `ci-gate` option for the expected head — filed as #690. - A test gap on the red-report harvest path — filed as #691. ## Testing - Regression tests in `core/test/PhasePrHarnessTest.scala`: stale red, stale green, a resume that pushed nothing new, a timeout while the forge keeps the old head, a first push, and an unreadable HEAD. - CI green on the phase PR (#689). ## Release notes See `project-management/issues/IW-607/release-notes.md` (Czech). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Closes #607

claude added 13 commits 2026-10-03 01:19:16 +00:00

workflow(dx): Create diagnostic analysis for IW-607 ... a1d2992bf4

phase-pr and kanon awaitCi can gate a stale head and use up a CI attempt

Artifacts:
- analysis.md
- issue.md

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

chore(IW-607): update review-state to tasks_ready 26b5467c78

workflow(dx): Create task index for IW-607 ... 113d9e6057

phase-pr and kanon awaitCi can gate a stale head and use up a CI attempt

Artifacts:
- tasks.md
- review-state.json (tasks_ready)

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

docs(IW-607): intake verdict — agent-ready ... 70c9d3d791

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Phase 1: context ready ... 7ca0dbe900

Issue: IW-607
Phase: 1 - A resumed phase-pr judges the head the forge held before its push

This commit marks the checkpoint for phase 1 context generation.
Use dx-fix IW-607 to return to this point.

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

chore(IW-607): update review-state to context_ready 2c0430973c

workflow(dx): Generate Phase 1 tasks for IW-607 ... 62f4361579

A resumed phase-pr judges the head the forge held before its push

Artifacts:
- phase-01-tasks.md

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

chore(IW-607): update review-state to tasks_ready d994710e77

Phase 1: A resumed phase-pr judges the head the forge held before its push (#689) d8e9c7ee6a

chore(IW-607): update review-state for phase 01 c70446e297

chore(IW-607): mark phase 01 complete in tasks.md 472ddf579c

chore(IW-607): file follow-ups as #690 and #691 ... c2634c37fb

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

docs(IW-607): add release notes ... b479768366

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

mph added 1 commit 2026-10-03 01:20:21 +00:00

chore(IW-607): update review-state to all_complete 7f7b132f9c

mph requested changes 2026-10-06 21:18:12 +00:00

mph left a comment

Copy link

We have conflicts.

We have conflicts.

mph added 1 commit 2026-10-06 21:38:45 +00:00

docs(IW-607): rework for final PR review 16 a3f154cc0f

mph added 5 commits 2026-10-06 22:16:39 +00:00

Merge remote-tracking branch 'origin/main' into IW-607 ... 86f0e611c4

# Conflicts:
#	core/commands/PhasePr.scala

chore(IW-607): move review-state card out of the tree 0999042a95

Phase 2: context ready ... 2002156f94

Issue: IW-607
Phase: 2 - Rework for final PR review 16

This commit marks the checkpoint for phase 2 context generation.
Use dx-fix IW-607 to return to this point.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

workflow(dx): Generate Phase 2 tasks for IW-607 ... 9ef0ec3b6d

Rework for final PR review 16

Artifacts:
- phase-02-tasks.md

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

test(IW-607): read the judged head from the forge's CI calls, not the card ... c490512408

IW-702 on main stopped every command from writing ci_gated_sha, so the
two stale-head tests that read it fail after the merge. They now assert
the head the gate read CI for.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

claude added 1 commit 2026-10-06 22:20:40 +00:00

Phase 2: Rework for final PR review 16 (#726) e90b50585c

mph added 1 commit 2026-10-06 22:21:59 +00:00

chore(IW-607): mark phase 02 complete in tasks.md 7279b59e3c

claude commented 2026-10-06 22:22:49 +00:00

Author

Member

Copy link

Implement: rework for review 16
The rework phase merged into IW-607 through #726; this pull request carries it.

Implement: rework for review 16 The rework phase merged into IW-607 through #726; this pull request carries it.

mph merged commit df28069261 into main 2026-10-07 07:31:11 +00:00

mph referenced this pull request from a commit 2026-10-07 07:31:13 +00:00

Merge pull request 'IW-607: A resumed phase-pr judges the pushed head, not a stale one' (#692) from IW-607 into main

Sign in to join this conversation.

Reviewers

No reviewers

mph

Labels

Clear labels   

agent-ready

  

bug

Something isn't working

  

contract

Run the tool-contract suite on this PR

  

documentation

Improvements or additions to documentation

  

duplicate

This issue or pull request already exists

  

enhancement

New feature or request

  

good first issue

Good for newcomers

  

help wanted

Extra attention is needed

  

invalid

This doesn't seem right

  

needs-human

  

needs-input

Intake asked the human a question on the issue; waiting for the answer

  

question

Further information is requested

  

wontfix

This will not be worked on

No labels

agent-ready

bug

contract

documentation

duplicate

enhancement

good first issue

help wanted

invalid

needs-human

needs-input

question

wontfix

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

2 participants

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

iterative-works/iw-cli!692

No description provided.