Codex Readiness Unit Tester FAQs

Question 1

Does this skill require an OpenAI API key or external internet access?

Accepted Answer

No, this skill is designed to run entirely locally. All LLM evaluations happen in-session within your current AI context, and deterministic checks use local filesystem scanning without external API calls.

Question 2

Where can I find the results of the readiness test?

Accepted Answer

Results are stored locally in a timestamped subdirectory within the .codex-readiness-unit-test/ folder, including report.json, report.html, and a summary.json file.

Question 3

How does the scoring system work?

Accepted Answer

Each check is weighted equally. A PASS contributes 100% of its weight, a WARN contributes 50%, and a FAIL or NOT_RUN contributes 0%. The overall project status is marked as FAIL if any individual check fails.

Question 4

Which project files does this skill analyze?

Accepted Answer

The skill primarily targets AGENTS.md and PLANS.md for content quality, while also checking for the existence of config.toml and specific repository structures required for Codex-compatible agents.

Question 5

What is the difference between Read-only and Execute modes?

Accepted Answer

Read-only mode collects evidence and performs documentation audits without running any scripts. Execute mode performs those same audits and then proceeds to run the actual build/test commands defined in your project to verify they work as described.

Codex Readiness Unit Tester

Codex Readiness Unit Tester

Características Principales

Casos de Uso

Características Principales

Casos de Uso