An exceptionally well-engineered operational skill: fully executable commands, explicit validation feedback loops, and thoughtful edge-case handling (conflicts, unattended mode, non-convergence). The only real cost is length — nearly all content earns its place, but a few rationale passages and the screenshots/EE sections could be tightened or split out.

Dimension	Reasoning	Score
Conciseness	The body is dense with non-obvious operational knowledge (review-round idempotency, upload-path collision 422s, EE symlink behavior) and assumes Claude's competence — no boilerplate explanation of what a PR is. A few justification passages ('a rebase rewrites the head SHA ... invalidating work you have already paid for') could be trimmed, matching anchor 4 'Efficient; minor instances of over-explanation' rather than the fully lean anchor 5.	4 / 5
Actionability	Guidance is fully executable throughout: copy-paste-ready `gh pr create` with heredoc body, a working base64/jq/`gh api` screenshot-upload script, `review-round.sh` invocation, and concrete conflict-resolution commands. This matches the anchor 'Fully executable; copy-paste ready code or commands; specific examples cover the common cases'.	5 / 5
Workflow Clarity	The process is explicitly sequenced (Execution Steps 1–10, Review rounds 1–3) with validation checkpoints and feedback loops: verdict judging with fix-commit-push-retry, void-round recovery via `/codex`, conflict-before-outage checking, the clean-round marker gating `gh pr ready`, and a stop-when-not-converging rule. This matches the anchor 'Clear sequence with explicit validation steps; feedback loops for error recovery'.	5 / 5
Progressive Disclosure	No bundle files exist, and external references (`docs/enterprise.md`, `frontend/AGENTS.md`, `REVIEW.md`, `.agents/skills/pr/review-round.sh`) are one level deep and clearly signaled with section names. Structure is good with well-organized sections, though the screenshots and EE companion sections are sizable inline blocks that could live in separate files — anchor 4 'Good structure; most content is appropriately placed' rather than the fully split level-5 pattern.	4 / 5
	Total	18 / 20 Passed

Description

78%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description with an explicit 'MUST use when' trigger and a clear two-action capability statement. Its main weakness is narrow capability coverage relative to what the skill body actually does, which keeps specificity at the mid anchor.

Suggestions

Broaden the capability statement with one or two more concrete actions the skill performs, e.g. 'fixes review findings, attaches frontend screenshots, and opens the EE companion PR' — this would lift specificity from anchor 3 toward 4.

Add a couple of natural trigger synonyms users might say, such as 'submit a PR' or 'request review', to strengthen trigger-term coverage toward the level-5 anchor.

Dimension	Reasoning	Score
Specificity	The description names the domain and two concrete actions ('Open a draft pull request on GitHub', 'drive CI review rounds until it is ready'), but coverage is not comprehensive — fixing review findings, screenshots, and the EE companion workflow are absent. It fits the anchor 'Names domain and 1-2 concrete actions, but not comprehensive' better than the level-4 anchor 'several specific actions'.	3 / 5
Completeness	It explicitly answers both questions: what ('Open a draft pull request on GitHub and drive CI review rounds until it is ready') and when ('MUST use when you want to create/open a PR') with a concrete trigger phrase. This matches the anchor 'Clearly and explicitly answers both what AND when with concrete trigger phrases'; the weaker level-4 anchor requires a 'when' that 'could be more explicit', which does not apply here.	5 / 5
Trigger Term Quality	Natural terms users would say are present ('pull request', 'PR', 'create/open a PR', 'CI review'), giving good keyword coverage. A few natural variations are missing (e.g. 'merge', 'submit for review'), so it sits at anchor 4 rather than 5, and it is clearly above anchor 3 ('some relevant keywords but missing common variations').	4 / 5
Distinctiveness Conflict Risk	'draft pull request on GitHub' and 'CI review rounds' carve a distinct niche with minimal conflict risk, though the bare 'PR' trigger could overlap with a PR-review or merge skill. Anchor 4 ('mostly distinct; minor overlap risk with closely related skills') is the best fit — not fully the level-5 clear-niche anchor given that adjacent-review-skill overlap.	4 / 5
	Total	16 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 missing	Warning

	Total	14 / 16 Passed

Repository: windmill-labs/windmill
Path: .agents/skills/pr/SKILL.md
Commit: af42c47

Reviewed: 8 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.