Agent skill

research-review

Iterative review of research documents using Rule of 5

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/research-review-charly-vibes-wai-2

SKILL.md

Iterative Research Review (Rule of 5)

Perform thorough research document review using the Rule of 5 - iterative refinement until convergence.

Setup

If research document path provided: Read the document completely

If no path: Ask for the research document path or list available research documents

Process

Perform 5 passes, each focusing on different aspects. After each pass (starting with pass 2), check for convergence.

PASS 1 - Accuracy & Sources

Focus on:

  • Claims backed by evidence
  • Source credibility and recency
  • Correct interpretation of sources
  • Factual accuracy of technical details
  • Version/date relevance (is information outdated?)
  • Code references are correct (file:line exist and match claim)

Output format:

PASS 1: Accuracy & Sources

Issues Found:

[ACC-001] [CRITICAL|HIGH|MEDIUM|LOW] - Section/Paragraph
Description: [What's inaccurate or unsourced]
Evidence: [Why this is problematic]
Recommendation: [How to fix with specific guidance]

[ACC-002] ...

What to look for:

  • "This works by..." without code reference
  • Claims about codebase without verification
  • Outdated information (library versions, deprecated APIs)
  • Misinterpretation of source code
  • Assumptions presented as facts

PASS 2 - Completeness & Scope

Focus on:

  • Missing important topics or considerations
  • Unanswered questions that should be addressed
  • Gaps in the analysis
  • Scope creep (irrelevant tangents)
  • Depth appropriate for the topic
  • All research questions answered

Prefix: COMP-001, COMP-002, etc.

What to look for:

  • Research question asked but not answered
  • Obvious related topics not explored
  • Shallow treatment of complex topics
  • Too much detail on tangential topics
  • "Further research needed" without followthrough

PASS 3 - Clarity & Structure

Focus on:

  • Logical flow and organization
  • Clear definitions of terms
  • Appropriate headings and sections
  • Readability for target audience
  • Jargon explained or avoided
  • Consistent terminology

Prefix: CLAR-001, CLAR-002, etc.

What to look for:

  • Jumping between topics without transitions
  • Technical terms used without definition
  • Conclusions before supporting evidence
  • Redundant sections
  • Confusing or ambiguous language

PASS 4 - Actionability & Conclusions

Focus on:

  • Clear takeaways and recommendations
  • Conclusions supported by the research
  • Practical applicability to the project
  • Trade-offs clearly articulated
  • Next steps identified
  • Decision-making guidance provided

Prefix: ACT-001, ACT-002, etc.

What to look for:

  • Research without recommendations
  • Conclusions that don't follow from findings
  • "Interesting but..." without actionable insight
  • Missing implementation guidance
  • No clear "what should we do?"

PASS 5 - Integration & Context

Focus on:

  • Alignment with existing research
  • Connections to specs and requirements
  • Relevance to current project goals
  • Contradictions with established decisions
  • Impact on existing plans
  • References to related work

Prefix: INT-001, INT-002, etc.

What to look for:

  • Contradicts previous research without acknowledgment
  • Ignores existing patterns in codebase
  • Doesn't reference related specs or docs
  • Recommendations conflict with project direction
  • Missing cross-references

Convergence Check

After each pass (starting with pass 2), report:

Convergence Check After Pass [N]:

1. New CRITICAL issues: [count]
2. Total new issues this pass: [count]
3. Total new issues previous pass: [count]
4. Estimated false positive rate: [percentage]

Status: [CONVERGED | ITERATE | NEEDS_HUMAN]

Convergence criteria:

  • CONVERGED: No new CRITICAL, <10% new issues vs previous pass, <20% false positives
  • ITERATE: Continue to next pass
  • NEEDS_HUMAN: Found blocking issues requiring human judgment

If CONVERGED before Pass 5: Stop and report final findings.

Final Report

After convergence or completing all passes:

## Research Review Final Report

**Research:** [path/to/research.md]

### Summary

Total Issues by Severity:
- CRITICAL: [count] - Must fix before using research
- HIGH: [count] - Should fix before using research
- MEDIUM: [count] - Consider addressing
- LOW: [count] - Nice to have

Convergence: Pass [N]

### Top 3 Most Critical Findings

1. [ACC-001] [Description] - Section [N]
   Impact: [Why this matters]
   Fix: [What to do]

2. [COMP-003] [Description] - Section [N]
   Impact: [Why this matters]
   Fix: [What to do]

3. [ACT-002] [Description] - Conclusions
   Impact: [Why this matters]
   Fix: [What to do]

### Recommended Revisions

1. [Action 1 - specific and actionable]
2. [Action 2 - specific and actionable]
3. [Action 3 - specific and actionable]

### Verdict

[READY | NEEDS_REVISION | NEEDS_MORE_RESEARCH]

**Rationale:** [1-2 sentences explaining the verdict]

### Research Quality Assessment

- **Accuracy**: [Excellent|Good|Fair|Poor] - [brief comment]
- **Completeness**: [Excellent|Good|Fair|Poor] - [brief comment]
- **Actionability**: [Excellent|Good|Fair|Poor] - [brief comment]
- **Clarity**: [Excellent|Good|Fair|Poor] - [brief comment]

Rules

  1. Be specific - Reference sections/paragraphs, provide file:line for code claims
  2. Verify claims - Actually check code references and factual statements
  3. Validate actionability - Research should drive decisions, not just inform
  4. Prioritize correctly:
    • CRITICAL: Factually wrong or misleading
    • HIGH: Significant gaps or unclear conclusions
    • MEDIUM: Could be clearer or more complete
    • LOW: Minor improvements
  5. If converged before pass 5 - Stop and report, don't continue needlessly

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results