feat(app-control): add /app-control:map refresh by kyle-sexton · Pull Request #9934 · melodic-software/claude-code-plugins

kyle-sexton · 2026-10-25T07:53:37Z

Summary

Adds the refresh action to /app-control:map, so the map's only writer also keeps a control skill current. It ports testing:refresh-feature-map steps 1-6 and their outcomes (clean, changed, blocked), and adds a source wave capped at 6, one live pass through /app-control:drive, and triage of each problem as doc drift, driver gap or product regression.

Fix

map refresh edits only the control skill's directory and commits on control-skill-upkeep/<slug>. An existing feature-map-upkeep/<slug> branch is reported, never adopted.
A product regression is reported and kept out of the upkeep PR. A legacy flat features/<f>.md map is migrated to a folder in the same upkeep PR.
upkeep-skip.sh and its test are copied into map/scripts/. No script-fingerprint verbs.
Eleven refresh cases and feature-map-maintainer-confines-edits are ported. Five cases are new: one per triage class, legacy-flat-map-migrates and source-wave-capped-at-six. One app-control changelog fragment (minor).

Verification

bash plugins/app-control/skills/map/scripts/upkeep-skip.test.sh exits 0. check-map.test.sh prints "37 cases, 0 failed". scripts/check-line-citations.py exits 0. /skill-quality:check map PASS (0 errors). /evals:validate exits 0.
grep -n 'control-skill-upkeep/' plugins/app-control/skills/map/SKILL.md matches line 185. grep -n -i fingerprint on the same file prints nothing.
claude plugin eval, Claude Code 2.1.310, 3 runs per arm, judge opus, no Bash, all VALID. New cases with the plugin vs without: doc drift 1.00/0.17, driver gap 1.00/0.00, product regression 1.00/0.00, legacy-flat-map-migrates 1.00/0.25, source-wave-capped-at-six 1.00/0.50. Ported feature-map-maintainer-confines-edits, run concurrently: before (testing@db62cbf3f) 1.00, after 1.00.
The eleven ported refresh cases (skill-creator evals.json, which claude plugin eval does not run) were run concurrently before and after with blind graders: suite mean 97.73 on both sides, no case lost or dropped. The evidence is in a PR comment.
Left out: Bash-granting eval runs (this host's eval sandbox refuses them), trigger-rate measurement.

…I and safety sections Map create discovers the repository over fixed rows before asking anything, asks only the residue (through /planning:interview when installed), writes the control skill and proves it once through /app-control:drive. The contract gains config layering, driver CLI and safety-floor sections. The 16 map-features cases and the unsafe-location case are ported, and six new cases cover discovery order, recipients, sign-in lockout, unattended stops, evidence survival and config layers. Refs: #9581 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…harpen the map eval cases Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…its graders Discovery stays read-only; the check that the app runs goes through /app-control:drive and its cleanup, and the no-orphan check covers every launch record drive wrote. The writer case can now fail its Write grader, the config and recipient regex graders match only what a correct answer says, case 12 keeps its refresh-skill expectation, and the run-/verify location rule states our reason in our words. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

… pin the control-skill location - --dir accepts only another .claude/skills/control-<slug> directory, the one place drive looks; real paths of the target, ignore file and pointer target must stay inside the repository. - The generated driver CLI applies the safety floors from config committed at HEAD and refuses a loosening from the local, environment or flag layers with exit 4; send, sign-in and write steps are CLI commands. Credentials always come from environment variables. - One sign-in rule first: any failure counts, one attempt per account per create run across every drive call, one capped rerun. No state file. - Explicit drive call sequence with the app slug and one --out per drive run outside the checkout; the no-orphan check matches pid, start time and command-line hash. - Hub last in the write order, resumable partial create, config/.gitignore checked with git check-ignore, unattended create is a dry run. - Hub table gains the app, driving tool and Attended runs only. rows. - Evals: id 12's always-true expectation replaced, the ported location case no longer forbids writing in its prompt, and the recipient and config cases grade the CLI as the enforcer. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…ecipient case's phrasing check The plugin description now names /app-control:map beside drive, with the catalog row regenerated. The recipient case's regex also accepts a negation pointing back at the allowlist ("isn't on it"), and its judge accepts an allowlist that stays empty until committed at HEAD; both edits were made after a first run (regex failed 3 of 3 on phrasing while the reply refused the unlisted recipient through the CLI). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…haracter limit Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…efaults are committed Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…9582-map-check

… map create Name the caller as the keeper of the per-run sign-in count, pass --unattended through to drive, launch nothing while a side-effect or shared-system row is open, recommend only safe values for safety questions, report discovered secrets by file and key, real-path check every written file, test the CLI's floor refusals in the proof, align the README and eval graders with the contract, and fix the upstream record's line citations. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Carry the untrusted-content framing spine verbatim, give the driver CLI a side-effect-free config command and test the floors through it with an exit-4-only pass, let the Floors check decide a resumed CLI, derive the drive selector from the resolved --dir slug, and skip the start check when drive would pick another app's sole control skill. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

check-map.sh reports a control skill's structural contract breaks as `<rule> | <path> | <detail>` lines (index to folders, sections, entry-point markers, links, shared/ use and why line, source globs), never edits, and exits 0 with findings. The map body's check action runs it, then judges copied recipes. Three eval cases cover a dead index line, a missing not-driven reason and a copied recipe, each with no file change. Refs: #9582 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…contract check-map.sh's shared-use and shared-why rules now point at the contract's "Shared items" section instead of being the only home for the convention. Refs: #9582 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

…9583-map-refresh

Refresh keeps a control skill in line with the app: skip check, a read-only source wave in waves of at most six, a live pass through /app-control:drive under the Ready/cleanup/evidence invariants, and triage into doc drift, driver gap and product regression. Changes go out on control-skill-upkeep/<skill slug> through one pull request; legacy testing feature maps and flat feature files convert in the same pull request. Ports upkeep-skip.sh and its test, the 11 refresh cases and the confines-edits case, and adds five new cases. Refs: #9583 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…ding .claude Inside an enclosing git repository, the git top level is not the project, so every index source glob looked dead. The default root is now the directory holding .claude/ (the git top level only for a directory outside .claude/skills/), the contract says sources are relative to it, and the no-shell fallback reports only what the rules find. The check evals now require the action's `<rule> | <path> | <detail>` line, which a run without the skill does not produce. Refs: #9582 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

map-legacy-flat-map-migrates seeded files into a repository with no commit, so the refresh correctly stopped as blocked at the skip check. The scaffold now initializes a repository when none exists and commits the seeded tree on main with a fixed author and date. Refs: #9583 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…tory map uses <slug> for the app slug, so `control-skill-upkeep/<skill slug>` read as control-skill-upkeep/clockwork. Name it control-skill-upkeep/control-<slug> everywhere, with the never-form beside the example, and move the cursor-pstack citation of the conversion section to its new lines. Refs: #9583 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…epository `git rev-parse --git-dir || git init` committed into an enclosing repository when the scaffold ran inside one. Init here unless this directory is already a repository's top level. Refs: #9583 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

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

Keep refresh in map's argument-hint with unspaced alternatives, and mark the literal backticks in the source-wave scaffold for shellcheck. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Main now holds map create (#9628) and map check (#9921) squashed. Their files take main's version; refresh composes onto main's SKILL.md, evals.json, README and the cursor-pstack record. The released map-create fragment and a sample for a grader main renamed are removed, and the cheat sheet is regenerated. Refs: #9583 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

kyle-sexton · 2026-10-25T08:07:09Z

Ported-case non-inferiority (skill-creator evals.json format, run the way Phase 4 ran it in #9628): testing:refresh-feature-map at db62cbf vs /app-control:map refresh at e8f3581, all 22 executors run concurrently, with-skill arm, each side graded on its own suite by blind fresh-context graders, no shell.

Before	After	Before %	After %
1 no-map-blocks	17 refresh-no-control-skill-blocks	100	100
2 skip-when-unchanged	18 refresh-skip-when-unchanged	100	100
3 unattended-chrome-refused	19 refresh-unattended-attended-only-refused	100	100
4 renamed-handle-corrected-in-map-only	20 refresh-renamed-handle-corrected-in-skill-only	75	75
5 lost-behavior-reported-not-mapped	21 refresh-lost-behavior-reported-not-mapped	100	100
6 open-pr-updated-not-duplicated	22 refresh-open-pr-updated-not-duplicated	100	100
7 clean-pass-records	23 refresh-clean-pass-records	100	100
8 driver-unavailable-writes-no-candidate	24 refresh-driver-unavailable-writes-no-candidate	100	100
9 working-tree-changed-during-source-pass	25 refresh-working-tree-changed-during-source-wave	100	100
10 untracked-map-committed-with-corrections	26 refresh-untracked-skill-committed-with-corrections	100	100
11 untracked-map-with-outside-change-blocked	27 refresh-untracked-skill-with-outside-change-blocked	100	100

Suite mean 97.73 before and after; no case lost, none dropped. Non-inferior under skill-criteria step 6.

Case 4/20 misses on both sides on the same expectation (edits the handle in the feature file): both runs held the edit behind the attended ask that expectation 3 requires. Graders flagged that the two expectations conflict on an attended pass. The expectations were left unchanged.
Cases 6/22, 7/23 and 10/26 were graded on the recorded commit, push or record decision on both sides, since the executors had no shell.
20 of 22 runs score at the ceiling, so this suite separates the two sides very little.

A reply that read only SKILL.md planned to edit cli/ and re-drive on an unattended pass, because the rule lived only in refresh.md's triage table. The hub now says an unattended pass reports the fix unwritten and drives nothing again. A flat feature file's conversion also updates its hub index line. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

"The send failure is a driver gap" matched no branch of the names-* graders. A branch for the class right after "is" or "as" covers it, while "is not a" still fails; pass and fail samples cover both. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

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

… blocked An unattended pass now reports a driver-gap fix unwritten. Ending it clean would record the pass and let the next one skip the unfixed gap, so it ends blocked and records nothing. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…asses class(?:es|ed) matches what classe[sd] matched; labelled becomes labeled. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

kyle-sexton · 2026-10-25T08:48:09Z

Local code and security reviews ran at ready, because review-venue read both CI lanes as absent. What happened to each finding:

Finding	Disposition	Basis
Sec 1: a fork PR with the upkeep branch name is taken as the upkeep PR	VALID (defer), #9937	Also on main: `plugins/testing/skills/refresh-feature-map/SKILL.md:177`
Sec 2: the source wave is read-only by prompt only, with no framing	VALID (defer), #9938	Also on main: `refresh-feature-map/SKILL.md:96-98`
Sec 3: an unattended driver-gap fix edits `cli/` (which enforces the floors) and re-drives	VALID (fix now)	An unattended pass reports the fix unwritten and ends blocked; case `map-triage-driver-gap-unattended` added
Sec 4: a refreshed control skill skips create's real-path checks	VALID (fix now)	Given and globbed directories are held to `naming.md#real-paths`, re-run before each write
Sec 5: `--unattended` now commits, pushes and opens a PR	No change	`refresh-feature-map` on main already runs unattended scheduled passes that open a PR; the approved plan ports it
Code C1: `upkeep-skip.sh` is copied in two plugins	No change in this PR	The copy is required (A4). Phase 8 (#9585) deletes the testing copy, so a sync check would be removed almost at once
Code I1: branch `control-skill-upkeep/control-<slug>` vs `<slug>`	INCORRECT	#9583 says `<slug>`; the branch is named for the control skill, one per skill, and the sanity grep matches
Code I2: `feature-map-upkeep/<slug>` cannot be derived	VALID (fix now)	Every `feature-map-upkeep/*` branch is listed and reported by name; case `map-reports-legacy-upkeep-branch` added
Code I3: the wave grader requires 6+2 but the skill allows any split	VALID (fix now)	The skill now fills each wave to the cap; the grader is unchanged
Code I4: the description overlaps `testing:refresh-feature-map`	VALID (fix now)	The phrase is dropped; a legacy map converts only when its path is passed
Code I5: flat control-skill conversion and multi-skill passes are untested	VALID (fix now)	Cases `map-converts-flat-control-skill` and `map-refreshes-each-control-skill` added
Code S1: a multi-skill pass stays on the first upkeep branch	VALID (fix now)	The pass returns to its starting branch after each skill
Code S2-S5: `## Next` shape, triage regexes, branch-grader boundary, fixture hub rows	VALID (fix now)	Fixed, with fail samples for the regexes

Evals at 726e4b6 (claude plugin eval, 3 runs per arm, no Bash, all VALID), scored with the plugin and without:

Case	With plugin	Without
map-triage-driver-gap	1.00	0.00
map-triage-driver-gap-unattended	1.00	0.00
map-triage-doc-drift	1.00	0.00
map-triage-product-regression	1.00	0.00
map-source-wave-capped-at-six	1.00	0.50
map-legacy-flat-map-migrates	0.92	0.25
map-reports-legacy-upkeep-branch	1.00	0.50
map-converts-flat-control-skill	1.00	0.33
map-refreshes-each-control-skill	1.00	0.00

The last commit, ce769c8, changes only the outcome rule for an unwritten driver gap; no case grades that outcome word.

chatgpt-codex-connector · 2026-10-25T08:48:40Z

Codex Review Summary

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

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-25T08:50:18.837806Z	`edf85d7`	Draft marked ready

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.

github-actions

claude-security-review: 1 inline finding(s) on edf85d76c8493c445a98621a7ae3e6b8b1825eaa; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

github-actions · 2026-10-25T08:50:09Z

claude-review reviewed 40451ac8f2ebdc2be2d776fb637bf5e39fd5967a: 0 finding(s), 0 as inline comments.

I found no issues in the incremental change, which covers the 8 files that changed since 3410a23. I read .claude-lane/incremental.diff, the PR context, and the current refresh.md. Nothing here runs: it is skill prose, an eval case and a fixture, so I did not run the eval or the scaffold script.

Dirty-map block and copy-based undo (plugins/app-control/skills/map/reference/refresh.md:87-107). The new text is internally consistent.
- The git status --porcelain -- "<legacy dir>" check runs before the planned-conversion ask, the upkeep-branch step 6.1 and any write. Only ?? lines are tolerated, so untracked notes in the map do not block it. Staged or unstaged edits to tracked files do.
- The copy to mktemp -d is taken after the branch step and before the first git mv, so it holds the pristine legacy directory.
- The undo steps cover the index and the working tree. The legacy files are put back from the copy, and the copy is deleted only after a clean undo or after step 6.3 commits.
Changelog fragment (.changes/app-control/feat-9583-map-refresh-29fdd303.md). It matches the new behavior: a dirty tracked map blocks, and the undo restores from a copy taken outside the checkout. The bump level is still minor.
Line citations (docs/upstream/cursor-pstack.md). I checked the ranges against the head file, and they now land on the right sections:
- 85-127 is the source wave.
- 129-150 is the live pass.
- 152-161 is the triage.
- 214-240 is Converting a legacy map.
New eval map-legacy-dirty-map-blocks.
- The scaffold commits a clean legacy map and then appends one uncommitted line to a tracked file, which matches the case's ls-style expectation of M .claude/skills/feature-map/start-a-timer.md.
- The slug control-clockwork-web follows from the fixture's Application line.
- The regex grader's \bblocked\b does not match "unblocked", as the fail sample shows. Together with the LLM grader, it can tell a refusal from a converted map.
- The prompt tells the model that git status prints the dirty line and that the decide step prints run, so the case does not depend on a shell.

No hold. The change spends no billed resources.

github-actions · 2026-10-25T08:50:09Z

claude-security-review reviewed 40451ac8f2ebdc2be2d776fb637bf5e39fd5967a: 0 finding(s), 0 as inline comments.

No security findings in the incremental change since 3410a23.

What changed

plugins/app-control/skills/map/reference/refresh.md is an instruction-surface edit. It adds two behaviours:
- A legacy map with uncommitted changes to tracked files now ends blocked before anything is written.
- The undo path now restores the legacy files from a copy taken outside the checkout, instead of from git restore --source=HEAD.
The changelog fragment and the docs/upstream/cursor-pstack.md row are prose and line-citation updates.
The new eval map-legacy-dirty-map-blocks has a fixture, a sample file and scaffold.sh.

Instruction-exception check

I did not read docs/conventions/instruction-exception-register/README.md, so the protected-class list was not consulted. I fell back to judging by consequence.
The diff removes no control. It narrows the old git restore --source=HEAD undo, which would overwrite uncommitted edits. The new dirty-map guard blocks before any write, so nothing can be overwritten.
The undo now deletes only paths the conversion listed, plus any directory the conversion created that is left empty. Those paths sit under .claude/skills/control-<slug>/, and the real-path checks still run before each write. I found no way in the diff to point the deletion at an unrelated path.
The backup is made with mktemp -d outside the checkout. It holds only the legacy map's own files. The report names its path, and it is deleted only after the undo finishes or step 6.3 commits.

Other files

scaffold.sh runs in an eval sandbox. It guards git init so it never commits into an enclosing repository, and it only touches files under its own directory.
The fixtures, case.yaml and the sample JSON are inert data.
No .github/standards/trusted-actors/**, check-trusted-trigger, CODEOWNERS or .gitmodules file changed, and no plugin manifest changed, so there are no manifest <field> lines.

critical-findings: 0

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: edf85d76c8

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Drive reads only control skills, so a legacy conversion's live pass had nothing to drive. The conversion is now written first, uncommitted on the upkeep branch, and both passes read the converted control skill. A pass that stops before the commit leaves those files in the working tree and names them. The legacy migration eval asks for the step order and judges it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…li/ steps The Config bullet called any write with no cli/ a driver gap, wider than the contract's "writes to an environment or external system", so a refresh read a click on the app's own scratch data as a gap and ended blocked. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

… specifies Plugin evals load no source-control plugin, and refresh step 6.4 then leaves the commit on the upkeep branch and reports the pull request as still to open. The grader and expected outcome now ask for that; a separate branch or pull request still fails. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

An entry already marked not driven is recorded state. Only a cli/ command the live pass found failing or missing, and left unwritten, ends the pass blocked. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions

claude-security-review: 1 inline finding(s) on 5e61ebd204c10f6bd818fef248cb40de9b78965d; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

github-actions

claude-review: 1 inline finding(s) on 5e61ebd204c10f6bd818fef248cb40de9b78965d; the summary is in this lane's pull request comment.

claude-lane: claude-review

…e pass stops An attended pass now shows the planned conversion and asks before the first write. A pass that stops before its commit moves the files back and returns to its starting branch, so the checkout is clean for the next run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…lat case The stricter wording failed a reply that stated step 6.4's plugin-absent branch as a condition. Either route on the one upkeep branch passes; a separate branch or pull request still fails. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions

claude-security-review: 1 inline finding(s) on 3410a23ab2d4fcf4c16c8d9785c9563b477324cb; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

github-actions

claude-review: 1 inline finding(s) on 3410a23ab2d4fcf4c16c8d9785c9563b477324cb; the summary is in this lane's pull request comment.

claude-lane: claude-review

…y map A refresh now refuses to convert a tracked legacy map with uncommitted changes, copies the legacy directory outside the checkout before the conversion's first write, and lists every path the conversion creates. The undo unstages and deletes those paths and copies the legacy files back from the copy instead of restoring from HEAD, so it works for an untracked map and leaves the tree as it was before the pass. A new eval case covers the dirty-map refusal. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

kyle-sexton and others added 26 commits October 24, 2026 02:37

fix(app-control): drop angle brackets from the map skill description

7f33131

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

fix(app-control): state the recipient floor in map's prove step and s…

9c69a22

…harpen the map eval cases Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Merge branch 'main' into feat/9581-map-create

5338727

fix(app-control): keep the plugin description under claude.ai's 500-c…

63378bf

…haracter limit Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

fix(app-control): mark create's send and write steps unproven until d…

8235d0d

…efaults are committed Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Merge remote-tracking branch 'origin/feat/9581-map-create' into feat/…

9187873

…9582-map-check

Merge branch 'main' into feat/9581-map-create

9033e10

test(app-control): cover the git top level as check-map's default root

794555d

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0177ybXguBnnGV4UTksrdzd9

Merge branch 'feat/9582-map-check' into feat/9583-map-refresh

0768d7a

Merge remote-tracking branch 'origin/feat/9581-map-create' into feat/…

e0bcf52

…9583-map-refresh

fix(app-control): write map's argument-hint with unspaced alternatives

67f6765

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

Merge feat/9582-map-check into feat/9583-map-refresh

0d8dea5

Keep refresh in map's argument-hint with unspaced alternatives, and mark the literal backticks in the source-wave scaffold for shellcheck. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Merge branch 'main' into feat/9583-map-refresh

faa636d

This was referenced Oct 25, 2026

fix(source-control): map upkeep can adopt a fork PR that shares the upkeep branch name #9937

Open

fix(app-control): enforce read-only on the map-upkeep source wave and frame its source text #9938

Open

kyle-sexton and others added 6 commits October 25, 2026 04:29

test(app-control): start the branch-handling scaffolds on main

1cd30ad

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

fix(app-control): follow map's moved line ranges in the pstack record

726e4b6

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

test(app-control): spell the triage regexes and one prompt so typos p…

edf85d7

…asses class(?:es|ed) matches what classe[sd] matched; labelled becomes labeled. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

kyle-sexton marked this pull request as ready for review October 25, 2026 08:48

github-actions Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md

chatgpt-codex-connector Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md

Comment thread plugins/app-control/skills/map/reference/refresh.md

kyle-sexton mentioned this pull request Oct 25, 2026

fix(app-control): run map upkeep passes on the tree the upkeep PR ships #9947

Open

kyle-sexton and others added 4 commits October 25, 2026 04:58

fix(app-control): block a refresh only on a driver gap the live pass hit

5e61ebd

An entry already marked not driven is recorded state. Only a cli/ command the live pass found failing or missing, and left unwritten, ends the pass blocked. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md Outdated

github-actions Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md Outdated

kyle-sexton and others added 2 commits October 25, 2026 05:49

github-actions Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md Outdated

github-actions Bot reviewed Oct 25, 2026

View reviewed changes

Comment thread plugins/app-control/skills/map/reference/refresh.md Outdated

kyle-sexton added this pull request to the merge queue Oct 25, 2026

Merged via the queue into main with commit 7321998 Oct 25, 2026
35 checks passed

kyle-sexton deleted the feat/9583-map-refresh branch October 25, 2026 14:26

kyle-sexton mentioned this pull request Oct 25, 2026

feat(testing)!: retire map-features and refresh-feature-map for /app-control:map #10033

Merged

Conversation

kyle-sexton commented Oct 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Fix

Verification

Related

Uh oh!

kyle-sexton commented Oct 25, 2026

Uh oh!

kyle-sexton commented Oct 25, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codex Review Summary

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot commented Oct 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Oct 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

kyle-sexton commented Oct 25, 2026 •

edited

Loading

chatgpt-codex-connector Bot commented Oct 25, 2026 •

edited

Loading

github-actions Bot commented Oct 25, 2026 •

edited

Loading

github-actions Bot commented Oct 25, 2026 •

edited

Loading