From wicked-garden
Run quality crew to validate and improve index accuracy to >=95%
npx claudepluginhub mikeparcewski/wicked-garden --plugin wicked-garden[--max-iterations N]search/# /wicked-garden:search:quality Run the Index Quality Crew - coordinated agents that validate brain index quality and automatically fill gaps until >=95% accuracy is achieved. ## Architecture ## Arguments - `--max-iterations` (optional): Maximum iterations before stopping (default: 10) ## Instructions 1. **Check current brain health and stats**: If brain is unavailable, inform the user and suggest starting it with `wicked-brain:server`. 2. **Run validation** to establish baseline: 3. If accuracy < 95%, iterate: **Scout phase**: Use Grep/Glob to discover patterns in ...
/qualityRuns comprehensive quality control: executes tests with coverage, checks code standards/docs/patterns, auto-fixes issues, produces score (0-100) and trends report.
/qualityRuns unified code quality checks including linting, complexity analysis, security scans, and dependency audits. Produces a scored report with issues, warnings, and recommendations. Supports subcommands like lint, fix, review.
/qualityGenerate quality intelligence report with tech debt analysis, health scoring, and predictive insights
/qualityRun comprehensive code quality checks including linting, static analysis, security scanning, and dependency audits
/qualityComprehensive quality assessment with multi-agent coordination
Run the Index Quality Crew - coordinated agents that validate brain index quality and automatically fill gaps until >=95% accuracy is achieved.
Scout Agent → Strategy Agent → Validator Agent → Executor
(explores) (plans) (approves) (runs)
--max-iterations (optional): Maximum iterations before stopping (default: 10)Check current brain health and stats:
curl -s -X POST http://localhost:4242/api \
-H "Content-Type: application/json" \
-d '{"action":"stats","params":{}}'
If brain is unavailable, inform the user and suggest starting it with wicked-brain:server.
Run validation to establish baseline:
/wicked-garden:search:validate
If accuracy < 95%, iterate:
Scout phase: Use Grep/Glob to discover patterns in the codebase that should be indexed:
@Service, @Entity, @Controller)models.Model, @Column).tsx)Strategy phase: Create extraction plan based on discoveries.
Execute phase: Re-ingest with the discovered content via wicked-brain:ingest.
Validate phase: Re-run /wicked-garden:search:validate to check improvement.
Stop when: >=95% accuracy achieved, OR plateau detected, OR max iterations reached.
/wicked-garden:search:quality
/wicked-garden:search:quality --max-iterations 3