| name | agent-readiness |
| description | Assess Gradient's readiness for governed agent work across docs, schemas,
profiles, harness, validation, evidence, and public-safe boundaries. Trigger:
/agent-readiness, /readiness.
|
| argument-hint | [--assess-only] [--fix] |
/agent-readiness
Readiness here is not framework setup. It is whether agents can safely move
Gradient work from intent to evidence without leaking private context or making
unsupported claims.
Pillars
- Public-safe boundary clarity.
- Lifecycle and module contract consistency.
- Profile/schema validation path.
- Evidence packet and feedback readiness.
- Harness skill/agent availability.
- Work tracker and closure mechanics.
- Fleet/context/policy contract maturity.
Current Baseline
Strong: architecture docs, decision log, profile examples, public-safe rules.
Weak: no automated gate, no evidence packet schema, no local Work adapter, no
Fleet run schema, no Context packet schema, no backlog lifecycle detector.
Gate
The load-bearing ship gate is ./scripts/validate.sh, plus reviewer
judgment for lifecycle semantics in docs/architecture.md and
docs/module-contracts.md when docs change.
Output
Score each pillar, name the highest-impact fix, and implement it only when the
scope is bounded and public-safe.
