Uh oh!
There was an error while loading. Please reload this page.
docs(skill): add audit-clerk-skill for drift detection - #156
Conversation
Stack: init-skills Part of a stacked PR chain. Do not merge manually. |
fd4e18e to
0ac582cCompare0ac582c to
93a54b9Compare97cadfd to
9255b9aCompare93a54b9 to
140f8d1Compare998207f to
71e2f9fCompare140f8d1 to
51a17d9Compare
|
51a17d9 to
b6a5d82Compare71e2f9f to
8078ef6Compareb6a5d82 to
6745d77Compare8078ef6 to
7a2025dCompare6745d77 to
ac7474cCompare7a2025d to
61ddd30Compareac7474c to
84a4ce1Compare84a4ce1 to
6f1b78fCompare61a2b63 to
b5936daComparee885c01 to
edbd7caCompareb5936da to
cc43a16Compareedbd7ca to
7ebc5d9Compare71bb4b1 to
2d03bbeComparea1a52fd to
6d0ef8aCompareAdds a user-invocable skill that cross-checks skills/clerk/ against the CLI source tree (packages/cli-core/src/commands/**) and proposes targeted edits when the two drift. Reads per-command README.md files as a secondary source, prefers --help over enumerating flags, and routes proposals between SKILL.md and references/*.md based on content type. Configured with effort: high and disable-model-invocation: true so it is a manual maintenance task, with an optional --apply argument to write changes rather than just proposing them.
2d03bbe to
20f1d14Compare📝 WalkthroughWalkthroughA new Claude skill definition file is added at Estimated code review effort🎯 2 (Simple) | ⏱️ ~12 minutes 🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. Comment |
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.claude/skills/audit-clerk-skill/SKILL.md:
- Line 112: There’s a contradiction between the rule text "No em-dashes" (the
forbidding statement) and the sample header that contains an em-dash (the
character “—”); resolve it by making them consistent: either remove/replace the
em-dash in the sample header (replace “—” with a hyphen or space) or relax the
rule to explicitly allow em-dashes and update the forbidding statement; update
the wording at the rule location (the "No em-dashes" line) and the sample header
text (the sample header containing “—”) so both reflect the chosen policy.
- Line 120: Add a language tag to the output-shape fenced code block that begins
with "# clerk skill audit — <YYYY-MM-DD>" so Markdown lint MD040 passes; edit
the fenced block delimiter (the triple backticks) to include a language
identifier (for example "markdown") immediately after the opening ``` to make
the fence look like ```markdown.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 1bc03b53-6268-4d3e-957b-c193da0c029b
📒 Files selected for processing (1)
.claude/skills/audit-clerk-skill/SKILL.md
| - **Never invent flags.** If a flag appears in a test but not in the command's argument parser, treat it as test-only and flag it for human review. | ||
| - **Preserve voice.** The existing skill is terse and third-person; match it. No first- or second-person drift. | ||
| - **No em-dashes** anywhere in proposals (repo style rule). |
There was a problem hiding this comment.
Resolve the em-dash rule contradiction in the output template.
Line 112 forbids em-dashes in proposals, but Line 121’s sample header uses one (—). This creates conflicting instructions for the same output.
Proposed fix
-# clerk skill audit — <YYYY-MM-DD>+# clerk skill audit - <YYYY-MM-DD>Also applies to: 121-121
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In @.claude/skills/audit-clerk-skill/SKILL.md at line 112, There’s a
contradiction between the rule text "No em-dashes" (the forbidding statement)
and the sample header that contains an em-dash (the character “—”); resolve it
by making them consistent: either remove/replace the em-dash in the sample
header (replace “—” with a hyphen or space) or relax the rule to explicitly
allow em-dashes and update the forbidding statement; update the wording at the
rule location (the "No em-dashes" line) and the sample header text (the sample
header containing “—”) so both reflect the chosen policy.
| Return the proposal as: | ||
| ``` |
There was a problem hiding this comment.
Add a language tag to the fenced code block.
The output-shape fence is missing a language identifier, which will fail markdown lint (MD040).
Proposed fix
-```+```markdown
# clerk skill audit — <YYYY-MM-DD>
...
-```+```🧰 Tools
🪛 markdownlint-cli2 (0.22.0)
[warning] 120-120: Fenced code blocks should have a language specified
(MD040, fenced-code-language)
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In @.claude/skills/audit-clerk-skill/SKILL.md at line 120, Add a language tag to
the output-shape fenced code block that begins with "# clerk skill audit —
<YYYY-MM-DD>" so Markdown lint MD040 passes; edit the fenced block delimiter
(the triple backticks) to include a language identifier (for example "markdown")
immediately after the opening ``` to make the fence look like ```markdown.
Uh oh!
There was an error while loading. Please reload this page.
Summary
New user-invocable skill that cross-checks
skills/clerk/againstpackages/cli-core/src/commands/**and proposes targeted edits when the two drift.Reads per-command
README.mdfiles as a secondary source (to flag mocked or hidden surfaces), prefers--helpover enumerating flags, and routes proposals betweenSKILL.md,references/agent-mode.md,references/auth.md, andreferences/recipes.mdbased on content type.Configured with
effort: high,disable-model-invocation: true(manual maintenance task), and an optional--applyargument to write changes rather than just proposing them.Stacked on #126.
Test plan
/audit-clerk-skilland confirm it produces the documented output shape (bucketed proposals, source citations, severity tags)--applyon a synthetic drift case and confirm it writes onlydriftandgapedits, leavingpolishsuggestions for review