The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the @openai/codex package using the npm package manager and the codex cask via Homebrew. These are standard procedures for deploying developer tools from established registries.
[PROMPT_INJECTION]: The instructions identify an indirect prompt injection surface related to the CLI tool's ability to process project-specific instructions from an AGENTS.md file.
Ingestion points: The AGENTS.md file located at the project root is read by the CLI at the start of every session.
Boundary markers: The skill documentation does not define specific delimiters or instructions for the agent to ignore potentially malicious embedded content within the AGENTS.md file.
Capability inventory: The CLI tool is capable of executing shell commands (e.g., npm run build, npm test, codex exec) and initiating network connections for authentication and MCP server communication.
Sanitization: The skill instructions do not include steps for sanitizing, validating, or escaping the content retrieved from the AGENTS.md file before it influences agent behavior.

codex-cli