---
name: "codex-project-readiness-check"
description: "Audit whether a repository is ready for productive Codex work"

metadata:
  tags: ["codex", "skill", "readiness", "audit"]
---

# Codex Project Readiness Check

Use this skill when the user needs: A repository readiness audit for productive Codex work. It checks project guidance, install scripts, tests, safety boundaries, verification commands, and workflow gaps so maintainers know what Codex can safely do next.

## Workflow

---
name: project-readiness
description: Audit whether a repository is ready for productive Codex work: setup, project guidance, tests, verification commands, safety boundaries, and maintainer workflow clarity.
user-invocable: true
trigger_keywords:
  - readiness
  - codex setup
  - project setup
  - audit this repo
  - prepare for codex
---

# Codex Project Readiness Check

Use this skill when the user wants to know whether a repository is ready for productive Codex-assisted development.

## Protocol

### 1. Inspect Project Entry Points

Read, when present:

- AGENTS.md
- README.md
- CONTRIBUTING.md
- package/build config
- test/lint/typecheck scripts
- CI workflows
- docs that describe architecture or local setup

### 2. Check Readiness Areas

Evaluate:

- Setup: Can a new agent install and run the project from documented commands?
- Guidance: Are project conventions, architecture, and boundaries explicit?
- Verification: Are test, lint, typecheck, build, and visual checks discoverable?
- Safety: Are secrets, protected files, destructive commands, and deployment actions clearly bounded?
- Workflow: Is there a clear review/PR/issue process?
- Agent Fit: Which tasks are safe for Codex now, and which need human supervision?

### 3. Produce Verdict

Use one of:

- READY - Codex can work productively with normal supervision.
- PARTIAL - useful, but missing important setup or verification pieces.
- NOT READY - Codex is likely to waste time or create unsafe changes.

### 4. Output

Return:

- readiness verdict,
- top 5 blockers,
- quick wins,
- recommended first Codex task,
- commands that should be documented or added,
- what a paid AI coding workflow audit would inspect next.

Keep the output direct and actionable. Do not modify files unless the user asks.

## Output

Return concise, actionable results and include generated files, commands, or next steps when useful.