CtrlK
BlogDocsLog inGet started
Tessl Logo

pr-review

Review an AI Factory pull request on the VPS runner and post one evidence-based verdict comment, reconciling prior review findings on subsequent runs.

58

Quality

66%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./.github/ai-review/pr-review/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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.

A lean, highly actionable instruction-only skill with concrete gh commands, an explicit output contract, and real validation checkpoints around the risky posting step. Its main gaps are minor: an unordered workflow, a rot-prone hard-coded default branch, and no use of reference files to shed detail from the main body.

DimensionReasoningScore

Conciseness

The body is a dense operational policy that assumes Claude's competence (no explanations of gh, CI, or PR basics), with nearly every line carrying an instruction. Minor deductions: the version-specific fact "The repository's default branch is `2.x`" will rot, and a few packed multi-clause sentences could be trimmed.

4 / 5

Actionability

Gives concrete commands (`gh pr view`, `gh pr diff`, `gh pr comment <number> --repo <owner/repo> --body-file <file>`), per-directory check lists, and an exact comment structure with named verdicts. Not a 5 because some commands are named without the flags needed to execute them and the checks are described rather than fully specified.

4 / 5

Workflow Clarity

Sections imply a clear sequence (context → fetch metadata → verify head SHA → review → compose → verify → post → return URL) with explicit checkpoints ("Immediately before posting, verify the PR is still open and its head matches the reviewed SHA") and a retry feedback loop ("inspect existing comments before retry to avoid duplicate comments"). Not a 5 because the ordering is never made explicit and some mid-workflow checkpoints are implicit.

4 / 5

Progressive Disclosure

A single-file skill (~80 lines) with no bundle files and well-organized sections (Context and scope, Project-specific checks, Publish the result). Structure is good, but the per-directory check details and the comment template could live in one-level-deep reference files, and there is no overview-to-reference navigation.

4 / 5

Total

16

/

20

Passed

Description

58%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 specific, action-oriented description that clearly states what the skill does, but it lacks any explicit "when to use" trigger guidance and misses natural synonym terms like "PR" and "code review". Adding a Use-when clause with those terms would materially improve it.

Suggestions

Add an explicit trigger clause, e.g. "Use when asked to review an AI Factory PR, code review a pull request, or re-check prior review findings on a PR."

Include natural synonyms users would say — "PR", "code review", "GitHub" — alongside "pull request" and "review".

Briefly name the verdict types (APPROVE / REQUEST_CHANGES / COMMENT) to sharpen the "what" toward comprehensiveness.

DimensionReasoningScore

Specificity

Names the domain ("AI Factory pull request on the VPS runner") and several concrete actions ("post one evidence-based verdict comment", "reconciling prior review findings on subsequent runs"), but coverage is not comprehensive — the nature of the review checks and verdict types are not enumerated.

4 / 5

Completeness

The "what" is clear (review the PR, post a verdict comment, reconcile prior findings), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3; "when" is only weakly implied by "on subsequent runs".

3 / 5

Trigger Term Quality

"pull request", "review", and "AI Factory" are relevant keywords, but common variations users would naturally say — "PR", "code review", "GitHub" — are absent.

3 / 5

Distinctiveness Conflict Risk

"AI Factory" and "VPS runner" carve a distinct niche, but generic phrasing like "Review a... pull request" means a user asking to review a PR could still trigger a generic code-review skill instead.

4 / 5

Total

14

/

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
lee-to/ai-factory
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.