feature-validate

I will help you validate a feature specification using AI-powered semantic analysis.

Usage:

• /feature-validate <feature-id> - Validate a specific feature spec
• /feature-validate --all - Validate all specs in artifacts/specs/
• /feature-validate <feature-id> --fix - Validate and suggest specific fixes

Instructions for me:

1. Load feature spec:

   • Read .mnemosyne/artifacts/specs/<feature-id>.md
   • If --all: Glob .mnemosyne/artifacts/specs/*.md
   • If not found: "Error: Feature spec '' not found"
   • Parse YAML frontmatter for spec metadata

2. Run DSPy validation:

   • Execute validation using optimized ReviewerModule v1:

     cd src/orchestration/dspy_modules
     uv run python3 specflow_integration.py ../../.mnemosyne/artifacts/specs/<feature-id>.md --json
     

   • Parse JSON output for validation results:
     • is_valid: Overall validation status (bool)
     • issues: List of specific problems found
     • suggestions: Actionable improvement recommendations
     • requirements: LLM-extracted requirements from spec
     • ambiguities: Detected vague terms and missing metrics
     • completeness_score: 0.0-1.0 quality score

3. Interpret results:

   • Excellent (score ≥ 0.9): "✓ Spec quality: Excellent"
   • Good (score ≥ 0.8): "✓ Spec quality: Good"
   • Fair (score ≥ 0.7): "⚠️ Spec quality: Fair - improvements recommended"
   • Poor (score < 0.7): "✗ Spec quality: Poor - significant issues found"

4. Display validation report:

   ✓ Validation complete
   
   Feature ID: <feature-id>
   Feature Name: <feature-name>
   Spec Location: .mnemosyne/artifacts/specs/<feature-id>.md
   Spec Version: <version>
   
   Validation Method: DSPy ReviewerModule v1 (semantic analysis)
   
   == QUALITY ASSESSMENT ==
   
   Completeness Score: <score>% (<rating>)
   Requirements Extracted: <count>
   Issues Found: <count>
   Ambiguities Detected: <count>
   Validation Status: <✓ Pass | ⚠️ Warning | ✗ Fail>
   
   == EXTRACTED REQUIREMENTS ==
   
   [First 5 requirements extracted by LLM:]
   1. <requirement>
   2. <requirement>
   3. <requirement>
   4. <requirement>
   5. <requirement>
   
   [If more than 5:]
   ... and <N> more requirements
   
   == ISSUES ==
   
   [If issues found:]
   ✗ <issue 1>
   ✗ <issue 2>
   ✗ <issue 3>
   
   [If no issues:]
   ✓ No issues detected
   
   == AMBIGUITIES ==
   
   [If ambiguities found:]
   🔍 <location>: <term>
      Question: <clarifying question>
      Impact: <why this matters>
   
   [If no ambiguities:]
   ✓ No ambiguities detected
   
   == SUGGESTIONS ==
   
   [If suggestions available:]
   💡 <suggestion 1>
   💡 <suggestion 2>
   💡 <suggestion 3>
   
   [If no suggestions:]
   ✓ Spec meets quality standards
   
   == NEXT STEPS ==
   
   [If score >= 0.8:]
   - Review spec: cat .mnemosyne/artifacts/specs/<feature-id>.md
   - Create implementation plan: /feature-plan <feature-id>
   
   [If score < 0.8:]
   - Address issues above (priority: high)
   - Clarify ambiguities: /feature-clarify <feature-id>
   - Re-validate: /feature-validate <feature-id>
   - After fixes, create plan: /feature-plan <feature-id>
   

5. Detailed fix suggestions (if --fix flag): For each issue/ambiguity, provide:

   • Location: Exact section/line in spec
   • Problem: What's wrong
   • Fix: Specific text to add/change
   • Example: Show before/after

   Format:

   == FIX #1: <issue summary> ==
   
   Location: <section> - <line range>
   Problem: <specific issue>
   
   Suggested Fix:
   Replace: "<current text>"
   With: "<improved text>"
   
   Example:
   Before: "API must be fast"
   After: "API must respond within 200ms (p95 latency) under normal load (1000 req/s)"
   
   Rationale: <why this fix improves spec quality>
   

6. Validation for --all flag:

   • Run validation on each spec sequentially
   • Display summary table:

     == BATCH VALIDATION RESULTS ==
     
     | Feature ID         | Score | Status   | Issues | Ambiguities |
     |--------------------|-------|----------|--------|-------------|
     | jwt-auth           | 92%   | ✓ Pass   | 0      | 0           |
     | api-rate-limiting  | 78%   | ⚠️ Warn  | 2      | 3           |
     | user-dashboard     | 65%   | ✗ Fail   | 5      | 7           |
     
     Summary:
     - Total Specs: 3
     - Passed (≥80%): 1
     - Warning (70-79%): 1
     - Failed (<70%): 1
     
     Average Score: 78%
     
     Recommended Actions:
     - Fix critical issues in: user-dashboard
     - Review and improve: api-rate-limiting