code-review-assist

Augments human code review by summarizing changes, surfacing key review questions, assessing test coverage, and identifying low-risk sections. Use when reviewing a diff, PR, or code snippet as a senior review partner.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

81%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 instruction skill with concrete response structures and explicit user-checkpoint gates. Its main costs are redundancy (duplicated severity-label definitions, overlapping concern bullets) and a monolithic single-file layout that inlines heuristics which could be progressively disclosed.

Suggestions

Define the conventional-comment severity labels (blocker/suggestion/nitpick/question) once — in Comment Format — and reference them from Key Concerns instead of repeating the full definitions in both sections.

Merge or tighten the overlapping 'Abstraction integrity' and 'Necessity' bullets, which both challenge interfaces added to serve a single implementation.

Move the detailed concern heuristics (mutation of shared state, premature optimization, abstraction integrity signals) into a references/ file and keep a short checklist in SKILL.md to reduce the always-loaded token footprint.

Dimension	Reasoning	Score
Conciseness	Mostly efficient skill-specific guidance, but the four severity labels (blocker/suggestion/nitpick/question) are defined in full twice — once under Key Concerns and again under Comment Format — and the Abstraction integrity and Necessity bullets overlap substantially, so the body could be tightened.	3 / 5
Actionability	Fully concrete instruction-only guidance: an eight-section response template, numbered session-planning steps, an explicit severity-label vocabulary with calibration rules, a named tracking file (review-session-notes.md), and a precise one-PR-at-a-time protocol.	5 / 5
Workflow Clarity	Multi-step processes are clearly sequenced with explicit checkpoints: complete one PR's full structure and "let the user respond" before proceeding, all draft comments "must be presented to the user for review before posting — no exceptions," and session coverage (reviewed/skipped/deferred) is reported at the end.	5 / 5
Progressive Disclosure	No bundle files exist and the single file is well-organized with clear section headers and a clearly signaled cross-reference to the vmcp-review skill, but at ~150 lines with detailed concern heuristics (mutation of shared state, abstraction integrity, premature optimization) inlined, some content could be split into a reference file — the sub-50-line exception for a reference-free skill does not apply.	4 / 5
	Total	17 / 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, concise, with an explicit Use-when clause and multiple concrete capabilities. Its only weaknesses are trigger-synonym coverage and not reflecting the full breadth of the body's capabilities.

Dimension	Reasoning	Score
Specificity	Names four concrete actions ("summarizing changes, surfacing key review questions, assessing test coverage, and identifying low-risk sections") but omits capabilities the body covers (drafting review comments, session planning/tracking), leaving minor gaps in coverage.	4 / 5
Completeness	Explicitly answers both what ("Augments human code review by summarizing changes, surfacing key review questions, assessing test coverage, and identifying low-risk sections") and when ("Use when reviewing a diff, PR, or code snippet as a senior review partner") with concrete trigger phrases.	5 / 5
Trigger Term Quality	"Use when reviewing a diff, PR, or code snippet" plus "human code review" are natural phrases developers would say, though common variants like "pull request" (spelled out) or "review my changes" are missing.	4 / 5
Distinctiveness Conflict Risk	The "senior review partner" framing carves out a human-in-the-loop niche distinct from autonomous reviewer agents, but the diff/PR/snippet triggers still overlap with generic code-review skills, leaving minor conflict risk.	4 / 5
	Total	17 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: stacklok/toolhive
Path: .claude/skills/code-review-assist/SKILL.md
Commit: 4c866b5

Reviewed: 7 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.