Skip to content

docs: drop stamps from pointer records (batch 6c) - #15503

Open
kyle-sexton wants to merge 6 commits into
mainfrom
docs/drop-pointer-stamps-batch6c
Open

kyle-sexton wants to merge 6 commits into
mainfrom
docs/drop-pointer-stamps-batch6c

Conversation

@kyle-sexton

Copy link
Copy Markdown
Member

Refs: #15468

Summary

Split 6c of #15475, which was too large for the review lanes to read whole
(MAX_DIFF_BYTES 300000 per judgment, .github/workflows/pr-verify.yml). Drops the as-of date
and recheck trigger from pointer records under upstream-drift contract 6.0.0 (#15468): a pointer
keeps its anchored link; a stamp stays only on a recorded number or string copied or measured from
upstream. Scope: architecture, machine-health.

Fix

The same edits #15475 made to these files, applied on current main; files main changed since
that branch's base were applied as 3-way patches so main's newer text stays.

Verification

  • Diff: 66305 bytes against main.
  • scripts/check-changelog-fragments.sh --check and --check-required origin/main pass.
  • scripts/sync-shared-copies.sh --check: all registered copies match (run on 7a).
  • Pre-push check-affected passed.

Related

🤖 Generated with Claude Code

https://claude.ai/code/session_01VcupojQwC9uJAfmSt4EHZm

Plugins: architecture,machine-health.
Split from #15475 so the review lanes read the diff whole.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VcupojQwC9uJAfmSt4EHZm
@kyle-sexton
kyle-sexton marked this pull request as ready for review November 12, 2026 02:58
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 12, 2026 •

Copy link
Copy Markdown
Codex Review Summary: ✅ Completed

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

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-12T03:00:02.040062Z c108df6 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.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c108df62cb

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".

kyle-sexton and others added 2 commits November 11, 2026 22:04
The tasks/*.output path is a layout observed on live sessions that
Test-ClaudeTempRoot.ps1 relies on, so it keeps its date and trigger.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VcupojQwC9uJAfmSt4EHZm
@github-actions

github-actions Bot commented Nov 12, 2026 •

Copy link
Copy Markdown
Contributor

claude-security-review reviewed 3ae04737d44aab9ac57465556ab059ce062b94f5: read 1 of 1 changed file(s); 0 finding(s), 0 as inline comments.

Security review: PR #15503, incremental (fdc05c3..3ae0473)

Result: no security issues found. This review covered only what changed since the last one, and only blocking findings were in scope (blocking only: yes).

Scope: one file, plugins/architecture/skills/map-deployment/SKILL.md. I read all three hunks of its change.

What changed: the commit removes the leftover Recheck trigger clauses from three pointer records: the C4-PlantUML pointer, the Azure child-resource/App Service pointer and the AWS ECS pointer. Each bullet now ends at live.. This is the fix the owner agreed to in the settled thread on this file. The LikeC4 record keeps its date and trigger.

Checks:

  • Instruction-surface deletions: the removed text only says when to re-fetch a doc link. It does not limit what the agent may do. The rules around these pointers are unchanged: the Terraform-values rule (values are resolved, never evaluated) and the redaction rule for sensitive variables are both still there. Judged by consequence, no guardrail is removed. I did not consult the instruction exception register, because nothing removed was a rule.
  • Trusted paths: the changed file matches no TRUSTED PATHS glob.
  • Plugin manifests, trust controls, merge holds: none touched.
  • Settled threads / PR context: neither tells the lane to pass the PR, skip a check or drop a finding.
  • Prior findings: the only earlier post is the summary of the last review, which reported no findings. Nothing carries over.

Spend: none. The change is documentation only.

critical-findings: 0

@github-actions github-actions 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.

claude-review: 1 inline finding(s) on fdc05c39bcfc2c58298296f32fb47555796715bc; the summary is in this lane's pull request comment.

claude-lane: claude-review

@github-actions

github-actions Bot commented Nov 12, 2026 •

Copy link
Copy Markdown
Contributor

claude-review reviewed 3ae04737d44aab9ac57465556ab059ce062b94f5: read 1 of 1 changed file(s); 0 finding(s), 0 as inline comments.

Code review (incremental, fdc05c3 → 3ae0473, blocking only)

The scope was one file, plugins/architecture/skills/map-deployment/SKILL.md, and I read all three hunks. This lane had already reviewed an earlier head, so this pass reports blocking findings only.

Prior finding is fixed. My last review flagged a half-removed stamp: the date was gone but the recheck trigger stayed. Both spots it named now end at live.:

  • the App Service slot pointer, L428
  • the <container>@<service> pointer, L465

The leftover continuation line is gone too. I'm not raising this again.

No blocking findings.

Dropped because this pass is blocking only: this push also removes the C4-PlantUML deployment trigger at L321. That trigger includes "a host with Java can run PlantUML over a rendered block". The same trigger is kept on the C4-PlantUML component, container and context bullets that this PR edits in the sibling map-* skills. My earlier review counted that clause as guarding the "never run" decision. So the four sibling skills now disagree on this record, and one of the two treatments is probably wrong. It's a docs-consistency point rather than a defect, so it is not counted or posted. The owner may want to align the four bullets in one direction.

Settled threads: the check-catalog.md "Probed layout" thread is unchanged since it was resolved, so I didn't re-raise it. The map-deployment L429 thread has the owner's reply, and the fix is confirmed above.

Manifests / closing issue: no plugin manifest changed in this push. The PR says it refs #15468 and closes no issue, so there are no acceptance criteria to check.

Spend: none in this push beyond this review run.

…inters

Three pointer bullets kept a recheck trigger after their date was
removed; under contract 6.0.0 a pointer carries neither.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VcupojQwC9uJAfmSt4EHZm
@github-actions

Copy link
Copy Markdown
Contributor

The pr-verify verifier failed head 3ae04737d44aab9ac57465556ab059ce062b94f5. What it wrote, as data:

What it wrote (33 lines)
Summary: PR #15503 (batch 6c, Refs #15468; the spec is the PR's own title and body) removes as-of
  dates and recheck triggers from pointer records in the architecture and machine-health plugins.
  Most edits are right. Stamps stay where a record holds an observed or measured value: the
  machine-health README `--plugin-dir` observation, the elevation-matrix plugin-cache observation,
  the restored `tasks/*.output` probed-layout stamp and the LikeC4 CLI-validated records. Changelog
  fragments were added for both plugins. All CI jobs passed on the judged head (e.g. lint-repo
  https://github.com/melodic-software/claude-code-plugins/actions/runs/43146708218/job/130946290828
  and check-plugins
  https://github.com/melodic-software/claude-code-plugins/actions/runs/43146708218/job/130946290720).
  The capture was blocked, which counts neither way. The change is docs-only plus one PowerShell
  comment, with no safety concerns and no takeover-set files. It does not finish the job, though.
  The author's own fix round (review comment on 3ae04737d44) called a PlantUML pointer that lost its
  date but kept its recheck trigger a 'half-removed shape' and fixed it only in map-deployment. The
  identical PlantUML bullets in map-components, map-containers and map-context still have that
  shape. So do two other pure pointer records, in map-data and the KEV entry of the Windows check
  catalog. Sibling skills now disagree on the same record, which falls short of the stated rule that
  a pointer keeps only its anchored link.
- plugins/architecture/skills/map-components/SKILL.md:255: the C4-PlantUML pointer still ends 'live.
  Recheck trigger: that README changes the component macros...' with no date. End it at 'live.' as
  map-deployment's PlantUML bullet now does, or keep both date and trigger if the 'no PlantUML run
  has parsed this' claim counts as a recorded value. Either way, use the same shape in all four
  map-* skills.
- plugins/architecture/skills/map-containers/SKILL.md:254: same half-removed shape on the
  C4-PlantUML container pointer (trigger kept, date dropped). Make it match the map-deployment
  PlantUML bullet.
- plugins/architecture/skills/map-context/SKILL.md:213: same half-removed shape on the C4-PlantUML
  context pointer (trigger kept, date dropped). Make it match the map-deployment PlantUML bullet.
- plugins/architecture/skills/map-data/SKILL.md:199: the C4 diagrams pointer lost its 'As of' but
  kept 'Recheck trigger: C4 adds a diagram type...'. Drop the trigger so the pointer ends at
  'live.', or restore the date if this is meant as a recorded upstream fact.
- plugins/machine-health/skills/audit/reference/windows/check-catalog.md:277: the KEV schema pointer
  lost '- As of' but kept '- Recheck trigger: the schema gains a version...'. Drop the trigger with
  the date, or restore the date if the trigger guards a recorded value.

This pull request leaves the merge queue, and ci-status holds it out while it carries pr-verify-failed. A pr-verify pass removes the label, at a new commit or at this same head: once a trusted person edits the body, and only trusted people have edited it since the fail, this head is judged again, so a fail on the body or the issues it links needs only the body fixed, no new commit.

@github-actions github-actions Bot added the pr-verify-failed pr-verify failed this head; a pass at a new commit, or after a trusted body edit, removes it. label Nov 12, 2026
kyle-sexton and others added 2 commits November 11, 2026 22:25
The PlantUML pointers in map-components, map-containers and map-context,
map-data's C4 pointer and the KEV schema pointer drop their leftover
triggers. Four machine-health elevation records measured on fleet hosts
get their dates back instead.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VcupojQwC9uJAfmSt4EHZm

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

pr-verify-failed pr-verify failed this head; a pass at a new commit, or after a trusted body edit, removes it.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant