agent-code-review-swarm
Agent skill for code-review-swarm - invoke with $agent-code-review-swarm
34
0%
Does it follow best practices?
Impact
98%
2.64xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./.agents/skills/agent-code-review-swarm/SKILL.mdQuality
Discovery
0%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This description is essentially a placeholder that provides no useful information about the skill's capabilities, triggers, or use cases. It only names the skill and provides an invocation command, which is insufficient for Claude to make informed decisions about when to select this skill from a pool of available options. This is among the weakest possible descriptions.
Suggestions
Add specific concrete actions the skill performs, e.g., 'Performs multi-perspective code reviews using a swarm of specialized agents to analyze code quality, security vulnerabilities, performance issues, and style compliance.'
Add an explicit 'Use when...' clause with natural trigger terms, e.g., 'Use when the user asks for a thorough code review, wants multiple perspectives on their code, mentions PR review, code quality analysis, or security audit.'
Remove the invocation syntax from the description (or move it to a separate field) and replace with substantive content that distinguishes this from a simple single-agent code review skill.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description provides no concrete actions whatsoever. It only says 'Agent skill for code-review-swarm' which is vague and abstract, giving no indication of what specific capabilities the skill provides. | 1 / 3 |
Completeness | The description fails to answer both 'what does this do' and 'when should Claude use it'. There is no explanation of capabilities and no 'Use when...' clause or equivalent trigger guidance. | 1 / 3 |
Trigger Term Quality | The only potentially relevant term is 'code-review' embedded in the skill name, but there are no natural keywords a user would say such as 'review my code', 'PR review', 'code quality', etc. The invocation syntax '$agent-code-review-swarm' is technical jargon, not a natural trigger. | 1 / 3 |
Distinctiveness Conflict Risk | While the name 'code-review-swarm' hints at a niche, the description itself is so generic that it provides no distinguishing information. It could conflict with any code review or code analysis skill. | 1 / 3 |
Total | 4 / 12 Passed |
Implementation
0%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is an aspirational feature catalog for a hypothetical 'ruv-swarm' tool rather than an actionable skill for Claude. It is extremely verbose, filled with non-executable pseudocode and descriptive JSON objects that list capabilities without teaching how to actually perform them. The lack of a coherent workflow, validation steps, and real executable commands makes it ineffective as a skill document.
Suggestions
Replace the fictional 'npx ruv-swarm' commands with actual executable steps using the declared tools (mcp__claude-flow__swarm_init, agent_spawn, task_orchestrate, gh CLI) so Claude can actually perform the described actions.
Reduce content to under 100 lines by removing the descriptive JSON catalogs of checks/metrics and instead providing a single concrete end-to-end workflow: receive PR number → gather diff → spawn review agents → collect results → post review.
Add explicit validation checkpoints and error handling, e.g., verify gh auth status, check that PR exists, validate agent responses before posting comments.
Move agent configuration details, comment templates, and CI/CD integration examples into separate referenced files to improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~400+ lines. Massive amounts of JSON config objects that merely list check names without actionable detail, redundant examples showing the same CLI pattern with minor variations, and sections like 'Best Practices' that state obvious things Claude already knows. Much of the content is aspirational documentation for a hypothetical tool rather than lean instructions. | 1 / 3 |
Actionability | The commands reference a non-standard tool ('npx ruv-swarm') with no installation or setup instructions, making none of the code executable. The JSON blocks listing checks (e.g., 'SQL injection vulnerabilities', 'Algorithm complexity') are descriptive catalogs, not actionable instructions. The bash scripts mix real gh CLI commands with fictional tool invocations, creating confusion about what actually works. | 1 / 3 |
Workflow Clarity | There is no clear sequential workflow with validation checkpoints. The skill presents a grab-bag of disconnected code blocks and configuration snippets without a coherent step-by-step process. No feedback loops or error recovery steps are defined despite the destructive nature of posting reviews and requesting changes on PRs. | 1 / 3 |
Progressive Disclosure | The content is a monolithic wall of text with no meaningful separation of concerns. References to 'swarm-pr.md' and 'workflow-automation.md' at the bottom are the only external links, but no bundle files exist to support them. Inline content that should be in separate reference files (agent configurations, YAML workflows, comment templates) bloats the main skill file enormously. | 1 / 3 |
Total | 4 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 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 |
Total | 10 / 11 Passed | |
- Repository
- ruvnet/claude-flow
- Commit
ca77f83
- 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.