CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

review-pr

Two-axis senior reviewer. Axis 1 (standards, "is it built right?") runs 6 specialized parallel sub-agents (Security, API/CRUD, Pattern enforcement, Simplicity, Reuse/DRY, Test hygiene) producing a severity-based report. Axis 2 (spec & behavior, "is it the right thing?") runs a 7th fresh-context sub-agent producing a behavior-change confirm-list for the human. Axes are never merged. Scoped to the current branch's diff against the default branch.

Quality

73%

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 ./.claude/skills/review-pr/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

77%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 highly actionable, well-sequenced review protocol with copy-paste commands, verbatim output templates, and explicit validation gates. Its weaknesses are verbosity — the axis-separation and mutation policies are each restated multiple times — and a monolithic structure with no progressive disclosure despite lengthy checklist content that would fit reference files.

Suggestions

State the axis-separation rule once in §3 and reference it from §5, §5b, and §6 instead of restating it ('never merged, co-ranked, or interleaved', 'never mix them into the C/H/M/L lists', etc.).

Consolidate the mutation-delta policy into a single section or a references/ file — Agent 6, Agent 7, and §5b each carry overlapping rules about survivors, scores, and the skip case.

Move the OWASP AST checklist and the behavior-surface table to references/ files so SKILL.md stays an overview and the detail is loaded only when the diff touches those surfaces.

DimensionReasoningScore

Conciseness

The body is mostly tight imperative instruction, but the axis-separation invariant is restated roughly six times ('Merging them destroys both', 'never merged, co-ranked, or interleaved', 'never merged with §5', 'never mix them into the C/H/M/L lists', 'The Axis-2 confirm-list is the single sanctioned exception'), and the mutation-delta policy is explained redundantly across Agent 6, Agent 7, and §5b. That places it between anchor 2 (several padded sections) and anchor 3 (mostly efficient, could be tightened); it earns 3 because the padding is concentrated in two repeated themes rather than spread throughout.

3 / 5

Actionability

Guidance is fully executable: exact commands ('git merge-base origin/<default-branch> HEAD', 'gh api repos/{owner}/{repo}/pulls/{number}/reviews'), verbatim output templates (the §5 summary table, the §5b confirm-list shapes, the 🧬 skip line), prescribed citation formats ('[file:line] <Mutator> survives — <what the mutant changed>, no test failed'), and a mandated closing sentence. Specific examples cover the common cases, including the stacked-PR and no-origin fallbacks.

5 / 5

Workflow Clarity

The protocol is explicitly sequenced (steps 0–7, with an 8-step branch scope lock) and includes real validation checkpoints and feedback loops: scope computed against a fetched remote tip, a mandatory user-confirmation gate ('Only after user confirmation, post ALL selected findings ... in a single API call'), and defined behavior for edge cases (stacked PR, no origin remote, no mutation adapter, marker line in the PR body). Checklists and severity categories structure the complex process.

5 / 5

Progressive Disclosure

There are no bundle files (no references/, scripts/, or assets/), and the 275-line body inlines content that clearly belongs in separate files: the ~10-bullet OWASP AST checklist, the behavior-surface table, and the duplicated mutation-delta policy would each fit a references/ file. Internal section structure is good (clear headers, numbered protocol), which keeps it above 2, but nothing is split out, which keeps it below 4.

3 / 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, third-person description that clearly conveys an unusual two-axis review mechanism with enumerated sub-agents and concrete outputs. Its main weakness is the absence of any 'Use when' trigger guidance and of the natural phrases ('code review', 'pull request') users would say, which caps completeness and weakens trigger quality.

Suggestions

Add an explicit trigger clause, e.g. 'Use when asked to review a PR, do a code review, or review the current branch's diff before merging' — without one, completeness is capped at 3.

Include the natural trigger terms users actually say — 'code review', 'pull request', 'PR review' — none of which currently appear in the description.

Trim internal process framing ('Axes are never merged') in favor of a user-facing capability statement, since the axis mechanics matter to execution, not to skill selection.

DimensionReasoningScore

Specificity

The description enumerates concrete actions: 'runs 6 specialized parallel sub-agents (Security, API/CRUD, Pattern enforcement, Simplicity, Reuse/DRY, Test hygiene) producing a severity-based report' and 'runs a 7th fresh-context sub-agent producing a behavior-change confirm-list'. It stays at 4 rather than 5 because the enumerated actions are internal process structure rather than comprehensive user-facing capability coverage, and it is above 3 because it lists several specific actions, not just 1-2. Voice is third person ('runs', 'produces'), so no voice penalty applies.

4 / 5

Completeness

The 'what' is clear (a two-axis review producing a severity report plus a behavior confirm-list, 'Scoped to the current branch's diff against the default branch'), but there is no 'Use when...' clause or equivalent explicit trigger guidance — the scoping sentence states what is reviewed, not when to invoke the skill. Per the judging guideline, a missing trigger clause caps completeness at 3.

3 / 5

Trigger Term Quality

Relevant domain keywords are present ('reviewer', 'Security', 'API/CRUD', 'Test hygiene', 'severity-based report', 'diff', 'branch'), but the natural phrases a user would actually say — 'code review', 'pull request', 'PR', 'review my changes' — never appear. It is above 2 (multiple relevant keywords, not one or two generic ones) but below 4 because the most common natural variations are missing.

3 / 5

Distinctiveness Conflict Risk

The two-axis structure, the enumerated sub-agent lenses, and the confirm-list output make it mostly distinct from adjacent skills ('Axes are never merged', 'confirm-list for the human'). It is not 5 because it claims the broad 'senior reviewer' niche with no distinguishing trigger phrase, leaving minor overlap risk with other code-review/PR-review skills.

4 / 5

Total

14

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 2 possibly missing (scripts/behavior-delta.sh, scripts/guards.config.sh)

Warning

Total

15

/

16

Passed

Repository

agranado2k/google-books-clojure

Path

.claude/skills/review-pr/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.