Runs a coding agent locally in the terminal, editing files and executing commands inside an OS sandbox with configurable approvals, plus headless `codex exec` for pipelines.
Skills
Sandboxed Code Execution
Edits files and runs shell commands inside an OS-level sandbox, escalating to the user for approval per the configured permission mode.
Headless Exec Runs
Executes prompts non-interactively through `codex exec` so coding tasks run inside CI jobs, scripts, and other repeatable workflows.
MCP Server Integration
Attaches local or remote MCP servers via `codex mcp` and exposes Codex itself as an MCP server for other agent clients.
Session Resume
Reopens a recent conversation from the current repository with `codex resume` and searches across stored local sessions.
AGENTS.md Project Config
Generates and reads AGENTS.md instructions through `/init` so repository-specific conventions steer the agent on every run.
Related Agents
SandBase Harness
Runs local-first, self-hosted AI agents with persistent sessions, sandboxed tools, memory, credentials, audit trails, r…
CrewAI
Multi-agent orchestration framework for building teams of AI agents that collaborate, delegate, and solve complex tasks…
Windmill
Turns scripts in Python, TypeScript, Go, and Bash into versioned APIs, workflows, and UIs, acting as an execution backe…
Claude Agent SDK
Lets Python and TypeScript developers embed the Claude Code agent loop in their apps, with built-in tools, hooks, subag…