Files
gh-bejranonda-llm-autonomou…/commands/analyze/quality.md
2025-11-29 18:00:50 +08:00

6.3 KiB

name, description, delegates-to
name description delegates-to
analyze:quality Run comprehensive quality control with autonomous auto-fixing autonomous-agent:orchestrator

Quality Check Command

🚨 CRITICAL: RESPONSE SAFETY REQUIREMENTS

SYSTEM-WIDE FAILURE PREVENTION: When generating ANY response content for this command, you MUST ensure:

  1. NEVER generate empty text blocks - All content blocks must have non-empty text
  2. NEVER use Unicode box characters (=, |, +, +, etc.) - Use safe ASCII alternatives
  3. ALWAYS provide fallback content for any section that might be empty
  4. VALIDATE all content blocks before finalizing response
  5. NEVER leave sections empty - Provide default values for missing data

SAFE RESPONSE PATTERN:

  • Use ASCII characters instead of Unicode box drawing
  • Ensure every score section has actual numeric values
  • Provide default content when data is missing
  • Never return empty strings or whitespace-only content
  • Always include actionable recommendations

FAILURE TO COMPLY: Will cause cache_control cannot be set for empty text blocks errors and break ALL Claude functionality.

Run comprehensive quality control check with autonomous fixing. This will:

  • Run all tests and analyze coverage
  • Check code standards compliance
  • Verify documentation completeness
  • Validate pattern adherence
  • Auto-fix issues when possible
  • Generate quality report with trends

How It Works

  1. Test Execution: Runs all tests and calculates coverage
  2. Standards Check: Validates code against style and standards
  3. Documentation Review: Checks for missing or incomplete docs
  4. Pattern Validation: Verifies adherence to learned patterns
  5. Auto-Fix Loop: Automatically fixes issues (repeats if needed)
  6. Quality Assessment: Calculates overall quality score (0-100)
  7. Trend Analysis: Compares against historical data

IMPORTANT: This command delegates to autonomous-agent:orchestrator which MUST present a detailed quality report to the user showing scores, test results, auto-fix actions, and specific recommendations. Silent completion is not acceptable.

Usage

/analyze:quality

Quality Scoring

  • Test Coverage: 30 points (aim for >80%)
  • Code Standards: 25 points (style, conventions)
  • Documentation: 20 points (completeness)
  • Pattern Adherence: 15 points (learned patterns)
  • Code Metrics: 10 points (complexity, maintainability)

Pass Threshold: 70/100

If score < 70, auto-fix loop is triggered automatically.

Example Output

The orchestrator MUST present results in this format:

============================================================
  QUALITY CHECK COMPLETED
============================================================

+-------------------------------------------------------+
| OVERALL QUALITY SCORE
+-------------------------------------------------------+
| Current Score: 88/100 [PASS]                         |
| Previous Score: 83/100                                |
| Trend: +5 points (improving)                          |
| Status: Above threshold (70+)                         |
+-------------------------------------------------------+

+-------------------------------------------------------+
| QUALITY BREAKDOWN
+-------------------------------------------------------+
| Tests (30 pts):         26/30 [PASS]                  |
|   45 passed, 2 failed | 88% coverage                 |
|                                                       |
| Standards (25 pts):     18/25 [WARN]                  |
|   18 style violations found                           |
|                                                       |
| Documentation (20 pts): 19/20 [PASS]                  |
|   97% of public APIs documented                       |
|                                                       |
| Patterns (15 pts):      15/15 [PASS]                  |
|   Adheres to 8/8 learned patterns                     |
|                                                       |
| Metrics (10 pts):       10/10 [PASS]                  |
|   Acceptable complexity levels                        |
+-------------------------------------------------------+

+-------------------------------------------------------+
| AUTO-FIX ACTIONS TAKEN
+-------------------------------------------------------+
| * Fixed 12 style violations (auto-formatted)         |
| * Added 3 missing docstrings                          |
| * Updated 1 outdated dependency                       |
| * Quality improved from 83 -> 88 (+5 points)          |
+-------------------------------------------------------+

+-------------------------------------------------------+
| REMAINING ISSUES
+-------------------------------------------------------+
| Tests:                                                |
| * test_user_login() - AssertionError (auth.py:45)     |
| * test_token_refresh() - Timeout (auth.py:89)         |
|                                                       |
| Standards:                                            |
| * 6 violations require manual review                  |
|   (complex refactoring needed)                        |
+-------------------------------------------------------+

+-------------------------------------------------------+
| RECOMMENDATIONS
+-------------------------------------------------------+
| 1. [HIGH] Fix 2 failing tests in auth module          |
|    -> Expected +4 quality points                      |
| 2. [MED] Refactor complex functions flagged           |
|    -> Expected +2 quality points                      |
| 3. [LOW] Review 6 manual style violations             |
|    -> Expected +1 quality point                       |
+-------------------------------------------------------+

Skills Loaded: quality-standards, testing-strategies
Agents Used: autonomous-agent:quality-controller, autonomous-agent:test-engineer
Auto-Fix Iterations: 2 (converged)
Patterns Stored: Quality pattern updated in .claude-patterns/
Check Time: 1.8 minutes

============================================================

Auto-Fix Details

When quality < 70, the plugin will:

  1. Run failing tests individually
  2. Fix style violations
  3. Generate missing documentation
  4. Suggest pattern improvements
  5. Re-check quality
  6. Iterate up to 3 times

See Also

  • /analyze:project - Autonomous project analysis
  • /learn:init - Initialize pattern learning database