agent-code-review-swarm

Agent skill for code-review-swarm - invoke with $agent-code-review-swarm

2.64x

Quality

—

Does it follow best practices?

Impact

98%

2.64x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Quality

Content

65%

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

The body is highly actionable, packed with executable gh and ruv-swarm commands, but it is verbose and monolithic with redundant sections and broken file references. Workflow sequences lack validation feedback loops for the destructive/outward-facing review operations.

Suggestions

Split the agent-specific checklists and config dumps into reference files under references/ and link to them from a concise overview, fixing the broken swarm-pr.md and workflow-automation.md links.

Add explicit validation checkpoints (e.g. parse/verify review output before approving or requesting changes) and a retry loop for failed gh API calls.

Consolidate the duplicate 'Specialized Review Agents' and 'Review Agents' sections to remove redundancy and reduce token cost.

Dimension	Reasoning	Score
Conciseness	The ~540-line body is dominated by concrete commands rather than concept explanations, but it is padded with redundant sections (the 'Specialized Review Agents' and 'Review Agents' blocks repeat each other) and large config/JSON dumps that could be tightened.	2 / 3
Actionability	It provides concrete, copy-paste-ready commands throughout ('gh pr view 123 --json files', 'npx ruv-swarm github review-security --pr 123 ...') with specific flags and a worked inline-comment posting loop.	3 / 3
Workflow Clarity	Sequences are present (fetch diff -> run review -> post comment), but workflows that approve, request changes, or post to PRs lack explicit validate->fix->retry checkpoints, which caps workflow clarity at 2 for these outward-facing operations.	2 / 3
Progressive Disclosure	Section organization exists, but the body is a monolithic single file with no bundle files; the two referenced files (swarm-pr.md, workflow-automation.md) do not exist on disk, so navigation is weak rather than well-signaled.	2 / 3
	Total	9 / 12 Passed

Description

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 frontmatter description is generic boilerplate that identifies the skill's name and invocation syntax but fails to convey its capabilities or when to use it. It lacks concrete actions, natural trigger terms, and any 'use when' guidance.

Suggestions

Rewrite the description to state concrete capabilities, e.g. 'Run specialized security, performance, and architecture review agents against a pull request and post findings via the GitHub CLI.'

Add an explicit 'Use when...' clause with natural trigger terms users would say, such as 'code review', 'review my PR', 'security review', or 'performance review'.

Drop the '$agent-code-review-swarm' invocation token from the description; it is internal syntax, not a capability or trigger.

Dimension	Reasoning	Score
Specificity	The description ('Agent skill for code-review-swarm - invoke with $agent-code-review-swarm') names a domain but states no concrete capability actions, matching the vague/abstract anchor rather than the 'names domain and some actions' level.	1 / 3
Completeness	It is auto-generated boilerplate that neither explains what the skill concretely does nor provides any 'Use when...' trigger guidance, so both what and when are very weak.	1 / 3
Trigger Term Quality	It offers only the internal invocation token '$agent-code-review-swarm' and the product name 'code-review-swarm'; no natural terms a user would actually say (e.g. 'code review', 'review my PR') are present.	1 / 3
Distinctiveness Conflict Risk	The 'code-review-swarm' niche is fairly specific so it is unlikely to collide with unrelated skills, but the boilerplate phrasing does not provide distinct natural triggers, keeping it below a 3.	2 / 3
	Total	5 / 12 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (543 lines); consider splitting into references/ and linking	Warning
relative_links	Relative link issues: 4 missing	Warning

	Total	14 / 16 Passed

Repository: ruvnet/claude-flow
Commit: 8ae8752

Reviewed: 2 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.