You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Split the Over-Length Sentences in the Repo Procedure Skills - #3587
Splits the over-length sentences in four repo-procedure skills so each reads zero for the sentence-length check, with no change to what any sentence requires.
Files:
.agents/skills/skill-lifecycle/SKILL.md
.agents/skills/audit-a-repo/SKILL.md
.agents/skills/standup-a-repo/SKILL.md
.agents/skills/resync-a-repo/SKILL.md
Their regenerated copies under .github/skills/ and .claude-plugin/fleet-skills/, plus the four .source-digests/ entries, produced by build_dist.py.
The exit-2 refusal list in skill-lifecycle is now a nested bullet list. Its items are the original items in the original order. One compound item ("one outside the root, ...") became a separate "a path outside the root, ..." item so the antecedent survives.
Every split was compared with its original, hunk by hunk. The drift classes were lost antecedent, dropped or half-attached reason, narrowed qualifier or citation, closed example list, and added claim. Local review rounds caught and fixed drift on new sentences. The last round found none.
The skill-lifecycle, audit-a-repo, standup-a-repo, and resync-a-repo skills
carried sentences over the 25-word limit. Each is split or restructured so
it keeps its original conditions, reasons, and scope. The exit-2 refusal
list in skill-lifecycle becomes a nested list with its items unchanged.
Generated copies and source digests are regenerated with build_dist.py.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The local review found splits that left a reason on only one half, widened
a refusal, or detached a condition from its clause. Each is reworded so the
original reach is kept.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The second local review found an either without an or, a rule limited to
one packaging shape that read as general, a lost antecedent, and a reason
and a citation covering only one half of a split sentence.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The third local review found an either-or reduced to an obligation, a
citation covering only the last of three sentences, an apposition that lost
its antecedent, and a verbless fragment.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The fourth local review found a section citation reaching only the first
of three converging steps, and a pronoun without an antecedent.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The fifth local review found six splits whose section citation moved onto
the last piece only. Each citation now reaches the pieces it covered.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The sixth local review found a causal link the original never stated and a
never-from-a-copy qualifier that moved off the instruction it governed.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The seventh local review found a section citation that no longer covered a
bullet's lead and a procedure reference that no longer covered the gate.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The eighth local review found that the shape the step names no longer
described both the summary and the reference files after the split.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A recorded local strict-review pass covers head 04c30136c91a9eea566dff6504d63ae84c294324, the content this pull request carries at that commit against develop, and it recorded 0 findings.
The reason will be displayed to describe this comment to others. Learn more.
🟡 Changes recommended
An active P2 nit remains in .agents/skills/resync-a-repo/SKILL.md, where the wording ambiguously associates the fetched RESYNC.md file with section 1 rather than clearly stating the required file relationship.
This pull request splits over-length sentences in four repository-procedure skills while preserving their operational requirements. It updates the canonical sources, regenerates the GitHub and Claude distributions, and refreshes the source digests.
Changes:
Splits long prose in skill-lifecycle, audit-a-repo, standup-a-repo, and resync-a-repo.
Restructures the skill-lifecycle refusal list into nested bullets while preserving its items and order.
Regenerates distributed skill copies and updates the four source-digest files.
Review findings remain in .agents/skills/audit-a-repo/SKILL.md and .agents/skills/resync-a-repo/SKILL.md, including the noted rationale, direct-push prohibition, and RESYNC.md file-reference issues.
File
Description
.github/skills/standup-a-repo/SKILL.md
Generated GitHub distribution of the standup skill.
.github/skills/skill-lifecycle/SKILL.md
Generated GitHub distribution of the lifecycle skill.
.github/skills/resync-a-repo/SKILL.md
Generated GitHub distribution of the resync skill.
Copilot found that the split left the file apposition attached to the
section number. The sentence now names section 1 of the file, then says the
file is fetched with the hub.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A recorded local strict-review pass covers head ac5c7ac796f52f4b2094cc3f01fff2fead4e1983, the content this pull request carries at that commit against develop, and it recorded 0 findings.
The reason will be displayed to describe this comment to others. Learn more.
🔵 Needs a closer look
A remaining sentence in .agents/skills/audit-a-repo/SKILL.md has an unclear antecedent and obscures the rationale for triggering on already-conformant repositories.
No open issues 🎉
1 previously missed issue
In code that hasn't changed since last review
Clarify the rationale for triggering on conformant repositories
.agents/skills/audit-a-repo/SKILL.md:4
[P2] Clarify the trigger rationale
The split leaves It is also because a repo that believes it is conformant is the case most often skipped without a clear antecedent for It, and it makes the repo itself read as the skipped case. This malformed sentence obscures why the skill triggers for an already-conformant repo. Keep the original meaning with a complete sentence such as This is the case most often skipped.
Copilot found that the repo, rather than the case of a repo believing it is
conformant, read as the case most often skipped. The sentence now names the
case explicitly.
Part of #3531.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A recorded local strict-review pass covers head 1c63b7f71347a4f8c69858b2e8f7be8943f36f29, the content this pull request carries at that commit against develop, and it recorded 0 findings.
Fixed in 1c63b7f: the finding on the trigger rationale in .agents/skills/audit-a-repo/SKILL.md line 4 is addressed. The sentence now names the case, a repo believing it is conformant, as the one most often skipped.
… and the Secrets Shell Library (#3589)
## Summary
Promotes `develop` to `main`, carrying the sentence-length splits of the
skill sources and the secrets shell library:
- #3582: splits the over-length sentences in the git and worktree
skills.
- #3584: splits them in the local review, agent conduct, and
check-this-repo skills.
- #3581: splits them in the comments of `lib/sh/secrets.sh`.
- #3585: splits them in the release skills.
- #3583: splits them in the codestyle and guard skills.
- #3586: splits them in the comment and doc style skill.
- #3587: splits them in the repo procedure skills.
The skill changes include their generated copies.
Closes#3577
This is part of #3531, which stays open.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Splits the over-length sentences in four repo-procedure skills so each reads zero for the
sentence-lengthcheck, with no change to what any sentence requires.Files:
.agents/skills/skill-lifecycle/SKILL.md.agents/skills/audit-a-repo/SKILL.md.agents/skills/standup-a-repo/SKILL.md.agents/skills/resync-a-repo/SKILL.md.github/skills/and.claude-plugin/fleet-skills/, plus the four.source-digests/entries, produced bybuild_dist.py.The exit-2 refusal list in
skill-lifecycleis now a nested bullet list. Its items are the original items in the original order. One compound item ("one outside the root, ...") became a separate "a path outside the root, ..." item so the antecedent survives.Every split was compared with its original, hunk by hunk. The drift classes were lost antecedent, dropped or half-attached reason, narrowed qualifier or citation, closed example list, and added claim. Local review rounds caught and fixed drift on new sentences. The last round found none.
Part of #3531.
🤖 Generated with Claude Code