github-code-review

Comprehensive GitHub code review with AI-powered swarm coordination

Quality

41%

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

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/github-code-review/SKILL.md

Quality

Content

38%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.

The content is an oversized monolith padded with repeated commands and benefit lists, scarred by a systematic '/'-to-'$' escaping bug that breaks several code and URL examples, and lacks validation checkpoints for its batch/destructive operations. Nothing is offloaded to bundle files despite the volume. Actionability and workflow clarity are moderate; conciseness and progressive disclosure are weak.

Suggestions

Fix the systematic escaping bug so closing tags read `</strong>`/`</summary>`/`</details>`, URLs read `https://cli.github.com/manual/`, and API paths read `repos/:owner/:repo/pulls/123/comments`.

Move the webhook handler, custom-agent JS, config YAMLs, and PR template into separate reference files under references/ and link to them, cutting SKILL.md to a lean overview.

Add explicit validate-then-retry checkpoints around destructive/batch steps (e.g. verify the diff parses before review-init, confirm comments posted before requesting changes, run validation before `--push-changes`).

Dimension	Reasoning	Score
Conciseness	The body is ~1140 lines for what is essentially one workflow, repeatedly re-stating `npx ruv-swarm github review-init` across Quick Start, Core Features, and five Example sections, plus padded ✅ 'Benefits' lists and full config/report dumps. It is noticeably verbose with several padded sections, though not pure concept explanation, placing it above the 1 anchor and at the 2 anchor.	2 / 5
Actionability	There is plenty of concrete bash (gh pr view/diff, npx ruv-swarm commands), but a systematic escaping bug replaces '/' with '$' throughout, breaking key examples (the inline-comment gh API call `repos/:owner/:repo$pulls/123$comments`, doc URLs like `https:/$cli.github.com$manual/`, and the custom-agent regexes `$app\.(get\|post...)` and `TODO$gi`). This yields 'some concrete guidance but incomplete / missing key details', not the mostly-executable 4 anchor.	3 / 5
Workflow Clarity	Sequences exist within sections, but the workflow involves batch and destructive operations (auto-merge, `--push-changes`, `--commit-fixes`, `review-comments --batch`, multi-PR) with no validate-then-retry feedback loops or verification checkpoints, so per the rubric cap workflow clarity cannot exceed 3. It is not a 2 because rough sequences and conditionals (e.g. grep for 'critical' before request-changes) are present.	3 / 5
Progressive Disclosure	No bundle files exist (references/, scripts/, assets/ are absent) and everything — webhook handler JS, custom-agent JS, config YAMLs, PR template, dashboards — is inlined into one monolithic 1140-line SKILL.md, matching the anchor 'content that clearly belongs in separate files is inlined'. The presence of a TOC and section headers keeps it just above the structureless 1 anchor.	2 / 5
	Total	10 / 20 Passed

Description

45%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.

The description is a single buzzword-laden sentence that states a domain but lists no concrete actions and provides no 'Use when' trigger guidance. It is distinguishable as a GitHub code-review skill yet risks overlapping with the many similar review skills. Specificity and trigger guidance are the main weaknesses.

Suggestions

Replace 'AI-powered swarm coordination' with concrete actions, e.g. 'Runs specialized security, performance, and architecture review agents on a PR and posts inline comments'.

Add an explicit trigger clause: 'Use when reviewing a GitHub pull request, running multi-agent code review, or automating PR review with ruv-swarm.'

Include natural trigger synonyms users actually say ('review my PR', 'code review on pull request', 'automated PR review') to improve trigger term quality and distinctiveness.

Dimension	Reasoning	Score
Specificity	The phrase 'Comprehensive GitHub code review' names the domain, but the only accompanying notion ('AI-powered swarm coordination') is a buzzword mechanism rather than a concrete action, matching the anchor 'Names the domain but actions are minimal or generic'. It is not a 3 because no 1-2 concrete review actions (e.g. 'post inline comments', 'request changes') are actually listed.	2 / 5
Completeness	A 'what' is present ('Comprehensive GitHub code review with AI-powered swarm coordination') but there is no 'Use when...' clause or equivalent trigger guidance, so per the rubric completeness is capped at 3 ('clear what, when missing or only weakly implied'). It is not a 4 because the 'when' is entirely absent, not merely imprecise.	3 / 5
Trigger Term Quality	'GitHub code review' is a reasonably natural phrase a user might say, but common variations ('review my PR', 'pull request review', 'review staged changes') are absent and 'AI-powered swarm coordination' is jargon rather than user speech. It sits above the generic 'Works with files' anchor (2) yet below the synonym-rich anchor (4).	3 / 5
Distinctiveness Conflict Risk	'GitHub code review' carves a recognizable niche, but 'code review' is a crowded space with several overlapping review skills, so it 'could still overlap with similar skills'. It is not a 4 because the description gives no distinctive trigger phrases that would disambiguate it from generic PR-review skills.	3 / 5
	Total	11 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (1141 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 2 missing, 3 suspicious	Warning

	Total	13 / 16 Passed

Repository: ruvnet/ruflo
Path: .agents/skills/github-code-review/SKILL.md
Commit: b14c79e

Reviewed: 3 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.