CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

skill-reviewer

Review skill PRs with structured severity-rated feedback covering token budgets, routing conflicts, and repo conventions. WHEN: "review skill", "review skill PR", "review skill changes", "check skill quality", "skill PR feedback".

Quality

87%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

86%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exemplary content structure: a lean overview that pushes all detail to real, one-level-deep, well-signaled reference files, with a sequenced workflow and an explicit validation step. The remaining improvements are small — an explicit feedback loop for failed validation and a concrete command for identifying changed files.

Suggestions

Add a feedback loop for the Validate step (e.g., 'If a suggested fix is not actionable, revise it before posting the review') so validation failures have explicit recovery guidance.

Make step 1 concrete (e.g., 'gh pr diff --name-only' or 'git diff --name-only main') instead of relying solely on the error-handling fallback for how to identify changed files.

DimensionReasoningScore

Conciseness

The 44-line body is lean throughout — a one-line purpose statement, five trigger bullets, a six-step workflow that links out to references instead of inlining them, and a two-row error table. It assumes Claude's competence (no concept explanations, no padded prose) and every token earns its place, matching the top anchor. Not a 4 because there is no over-explanation to trim; the only repetition (the References list) serves navigation.

5 / 5

Actionability

Each workflow step names its concrete artifact and source ("Run each category from the review checklist", "Assign severity per the severity guide", "Write the review per the output format"), and the error table gives a concrete fallback ("Estimate at ~4 chars per token") — mostly executable guidance with minor gaps, matching anchor 4. Not a 5 because step 1 ("Identify all changed skill files") gives no command or method, deferring entirely to the fallback error row; not a 3 because nothing is pseudocode or abstract.

4 / 5

Workflow Clarity

The six steps are clearly sequenced and end with an explicit validation checkpoint ("Validate — Verify suggested fixes are actionable with accurate file/line references"), and the error-handling table covers two failure modes — matching anchor 4's 'clear sequence with most checkpoints present; minor validation gaps'. Not a 5 because there is no feedback loop for what to do when step 6 validation fails (the error table only covers collection/token-counting failures); not a 3 because a validation checkpoint is explicitly present.

4 / 5

Progressive Disclosure

The body is a pure overview: all detail lives in four references, every one of which exists in references/, is linked with explicit markdown links in both the workflow and a References section, and is exactly one level deep (no nested pointers). All bundle files are linked and none are orphaned, matching the 'clear overview with well-signaled one-level-deep references' anchor. Not a 4 because there are no organization gaps — each reference maps to a distinct workflow step.

5 / 5

Total

18

/

20

Passed

Description

83%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: third-person, concrete about what the review covers, and equipped with an explicit WHEN clause of five natural trigger phrases. The only refinements available are minor — a few missing synonyms (e.g., "audit") and single-verb phrasing of the capability list.

DimensionReasoningScore

Specificity

"Review skill PRs with structured severity-rated feedback covering token budgets, routing conflicts, and repo conventions" names the domain and several concrete coverage areas (severity-rated feedback, token budgets, routing conflicts, repo conventions), matching the 'several specific actions; minor gaps' anchor. Not a 5 because it centers on a single verb ('Review') with coverage modifiers rather than a comprehensive list of distinct actions, and not a 3 because far more than 1-2 concrete actions are named.

4 / 5

Completeness

Both questions are answered explicitly: the "what" ("Review skill PRs with structured severity-rated feedback covering token budgets, routing conflicts, and repo conventions") and the "when" (WHEN: with five concrete trigger phrases), matching the top anchor. It is not a 4 because the trigger guidance is explicit and phrase-level, not merely implied or generic.

5 / 5

Trigger Term Quality

The WHEN clause quotes five natural user phrases ("review skill", "review skill PR", "review skill changes", "check skill quality", "skill PR feedback") with good variation, matching the 'good keyword coverage; a few natural terms missing' anchor. Not a 5 because common synonyms like "audit a skill" or "skill review" are absent; not a 3 because the listed phrases are exactly what a user would naturally say.

4 / 5

Distinctiveness Conflict Risk

The "skill PR" niche is well-delineated and every trigger phrase is skill-specific, so overlap risk is low but not minimal — a generic request like "review this PR" for a PR that happens to contain a skill has no matching trigger and could route to a general code-review skill, matching the 'mostly distinct; minor overlap risk' anchor. Not a 5 because that generic-PR gap remains; not a 3 because the triggers would not fire for ordinary document or code work.

4 / 5

Total

17

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

microsoft/GitHub-Copilot-for-Azure

Path

.github/skills/skill-reviewer/SKILL.md

Reviewed

Table of Contents

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.