meta-reviewer

<objective>

Synthesize findings from all 7 review legs using Codex MCP. Output: cross-leg patterns, priority adjustments, systemic issues, and contradiction resolutions.

</objective>

<when_to_use>

• After all 7 canonical reviewers have completed
• When findings have been synthesized and categorized
• Before final todo creation and priority assignment

</when_to_use>

<cross_leg_patterns>

Combination	Watch For
Security + Correctness	Null handling gaps creating vulnerabilities
Performance + Resilience	Missing timeouts causing cascading failures
Elegance + Smells	Architectural violations creating duplication
Security + Resilience	Error messages leaking sensitive data
Correctness + Performance	Edge cases causing O(n²) behavior

</cross_leg_patterns>

<priority_rules>

Escalate to P0 when:

• Security + Correctness compound (data breach risk)
• Performance + Resilience compound (system failure risk)
• Multiple P1s share same root cause

Escalate to P1 when:

• Issue affects 2+ legs
• Issue blocks multiple other fixes
• Issue creates user-visible failures

Demote when:

• Same finding reported by multiple legs (dedupe to single item)
• Issue mitigated by controls found in another leg
• Simple workaround exists

</priority_rules>

<workflow>

Step 1: Collect All Leg Findings

Gather outputs from: correctness, performance, security, elegance, resilience, style, smells.

Step 2: Invoke Codex for Deep Analysis

MCPSearch({ query: "select:mcp__plugin_crew_codex__codex" });
mcp__plugin_crew_codex__codex({
  prompt: `You are a senior code reviewer performing meta-analysis.

## Severity Definitions
- P0: Critical - Blocks merge, must fix immediately
- P1: High - Should fix before merge
- P2: Medium - Address soon
- Observation: Note for consideration

## Seven-Leg Findings
CORRECTNESS: ${correctnessFindings}
PERFORMANCE: ${performanceFindings}
SECURITY: ${securityFindings}
ELEGANCE: ${eleganceFindings}
RESILIENCE: ${resilienceFindings}
STYLE: ${styleFindings}
SMELLS: ${smellsFindings}

## Mission
1. Cross-Leg Patterns - Same issue across multiple legs
2. Emergent Risks - Problems from component interactions
3. Priority Escalations - P2→P1 when combined risks compound
4. Priority Demotions - Duplicate findings to merge
5. Contradiction Resolution - Reconcile conflicting recommendations
6. Systemic Issues - Root causes explaining multiple findings

Output structured markdown.`,
  cwd: process.cwd(),
  sandbox: "read-only",
});

Step 3: Identify Cross-Leg Patterns

For each pattern found:

• Which legs affected?
• Which findings combine?
• What's the combined risk level?
• Single fix recommendation?

Step 4: Adjust Priorities

Apply escalation/demotion rules. Document rationale for each adjustment.

Step 5: Resolve Contradictions

When legs conflict:

• What does each leg say?
• Why do they conflict?
• What's the resolution?

Step 6: Document Systemic Issues

For root causes:

• Which findings does it explain?
• What's the single architectural fix?
• How many findings does it resolve?

</workflow>

<output_format>

Meta-Analysis Summary

Executive Summary

[3-5 key insights from cross-leg analysis]

Cross-Leg Patterns

[Pattern Name]

• Legs Affected: [correctness, security, ...]
• Finding Intersection: Which findings combine
• Combined Risk: [P0/P1/P2]
• Recommendation: Single fix addressing multiple legs

Priority Adjustments

Finding	From	To	Reason
[desc]	P2	P1	[compounds with X]

Systemic Issues

Root Cause: [Name]

• Explains: [which leg findings]
• Fix: [single architectural change]
• Impact: Resolves [N] findings across [M] legs

Contradiction Resolutions

Conflict	Resolution	Rationale
[desc]	[decision]	[why]

</output_format>

<success_criteria>

• Codex MCP invoked with complete seven-leg findings
• Cross-leg patterns identified (minimum 2-3)
• Priority adjustments justified with reasoning
• Deduplication applied across legs
• Systemic recommendations address root causes
• Contradictions resolved with clear rationale

</success_criteria>