skill
Assess and improve codebase readiness for AI coding agents across style, tests, docs, architecture, CI, observability, security, and dev setup. Produces a scored report, prioritizes remediation, then executes the highest-impact fixes. Use when: "agent readiness", "is this codebase agent-ready", "readiness report", "make this codebase agent-friendly", "agent-ready assessment", "readiness audit", "prepare for agents". Trigger: /agent-readiness, /readiness.
/agent-readiness
Source: skills/agent-readiness/SKILL.md
Assess and improve codebase readiness for AI coding agents across style, tests, docs, architecture, CI, observability, security, and dev setup. Produces a scored report, prioritizes remediation, then executes the highest-impact fixes. Use when: "agent readiness", "is this codebase agent-ready", "readiness report", "make this codebase agent-friendly", "agent-ready assessment", "readiness audit", "prepare for agents". Trigger: /agent-readiness, /readiness.
Support primitive in the Harness Kit operating loop.
This generated excerpt gives readers the beginning of the live primitive contract before they jump to GitHub.
Assess how well this codebase supports autonomous AI coding agents.
Then fix the highest-impact gaps.
**Target:** $ARGUMENTS
## Execution Stance
You are the executive orchestrator.
- Keep prioritization, remediation approval, and final readiness judgment on the lead model.
- Delegate pillar assessments and bounded fixes to focused subagents.
- Use parallel fanout by default for independent pillars.
## Delegation Floor
Delegation floor applies: probe the roster first; dispatch two or more
providers for substantive work; direct solo only for mechanical, emergency,
user-forbidden, or fewer-than-two-providers cases. See
`harnesses/shared/AGENTS.md` (Roster).
Local lane guidance: Use specialized assessor lanes per readiness pillar and
separate remediation-builder lanes for approved fixes. Split by pillar or
ecosystem, prefer provider diversity for architecture/security/CI judgment, and
record waivers when profile CRUD is purely mechanical.
## Core Insight
The agent is not broken. The environment is. A codebase with fast feedback
loops and clear instructions makes any agent dramatically more effective.
A codebase with poor feedback loops defeats any agent you throw at it.
## Profile Contract
Assessment mode scores the repo. Profile mode manages the durable SDLC
contract at `.harness-kit/agent-readiness.yaml`. Route `profile create`,
`profile read`, `profile update`, `profile delete`, and `profile validate` to
`scripts/profile-crud.py`; the schema lives in
`references/profile-schema.yaml`. Missing profiles degrade to assessment-only
behavior, but stale, expired, blank, or placeholder-only waivers are invalid
once a profile exists.
## Workflow
### Phase 0: Profile CRUD
If the user asks to create, read, update, delete, or validate the readiness
profile, run the deterministic profile script first and stop after reporting
...