Getting started with Codex
- Source
- youtube.com
- Author
- OpenAI
- Date
Context
A coding can check its own work more effectively when the project tells it how. In OpenAI's Codex workshop, the presenters recommend putting build and test commands in AGENTS.md, a project instruction file loaded into the agent's . That makes the checks a developer would otherwise run afterward available during the agent's work. Watch the workshop.
They recommend keeping this file brief and pointing to separate documents for specialized tasks. When Codex repeatedly struggles to discover a command or project convention, recording the answer gives subsequent sessions a useful starting point. The file becomes maintained project guidance that helps the agent act and verify results. See the instruction-file discussion.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- MCP — The Model Context Protocol — an open standard that lets any AI assistant plug into any tool or data source without custom integration code.
Checking sign-in…
Loading comments…





