mycooperate-skill-lifecycle-pipelinelisted
Install: claude install-skill goondocks-co/myco
# Operating the Myco Skill Lifecycle Pipeline
Hands-on reference for managing skill candidates, skill records, and SKILL.md files using Myco's agent tools and Skills dashboard.
## Pipeline Architecture and Scheduling
**Asymmetric scheduling defaults** reflect a deliberate progressive disclosure model:
- `skill-survey` is `enabled: true` (auto-runs during idle, discovers candidates without user intervention)
- `skill-generate` and `skill-evolve` are `enabled: false` (require explicit opt-in)
This prevents runaway generation costs while ensuring candidate discovery happens automatically.
**Survey quality transformation:** The survey pipeline now runs cluster-first (grouping related spores before proposing candidates), producing 5 domain-scoped candidates vs the old pipeline's 102 candidates with 74.5% dismissed. This dramatically improves signal-to-noise ratio.
## Tool Reference
### vault_skill_candidates
Manages skill candidates — knowledge clusters identified as potential skills.
```
// List by status
vault_skill_candidates(action: "list", status: "identified"|"approved"|"dismissed"|"generated")
// Get single candidate
vault_skill_candidates(action: "get", id: "<uuid>")
// Approve a candidate
vault_skill_candidates(action: "update", id: "<uuid>", status: "approved", rationale: "...")
// Create a candidate manually
vault_skill_candidates(action: "create", topic: "...", rationale: "...")
```
Status lifecycle: `identified` → `approved` → `generated`. Dismiss with `st