What problem does it solve?
Building or running any app, script, CLI, or generator that calls the OpenAI API requires a valid OPENAI_API_KEY, and handling that credential incorrectly risks leaking plaintext secrets into logs, chat output, or tracked files. This Skill acts as a credential gate that inspects for existing keys without printing them, asks whether to reuse or create a key before any implementation work, and provisions new keys through an encrypted flow that never exposes plaintext.
Core Features & Use Cases
- Safe credential inspection: Checks the environment and local env files (.env, .env.local) for a usable OPENAI_API_KEY using silent exit-status checks that never print secret values.
- Secure key creation: Uses the OpenAI Platform connector picker and a public-JWK-only encrypted creation flow, with a local helper script that decrypts and writes the key only to a confirmed, in-workspace env file.
- Credential decision gate: Stops all implementation, scaffolding, and smoke testing until the user explicitly chooses reuse-existing-key or create-new-key.
- Use Case: A user asks to build an app that generates jokes from two input fields using AI. The Skill activates even though no key was mentioned, detects an existing key without revealing it, asks whether to reuse it or create a new one, and only then hands off to the implementation skill.
Quick Start
Ask Codex to build an app or script that uses the OpenAI API, and it will check your credentials and ask whether to reuse an existing key or create a new one before writing any code.
