Documentation
llx docs
How llx works at your terminal, in scripts and CI, and for other agents: the command line, the GitHub Action and its recipes, the seven tools, the modes, and the exit codes and events a pipeline relies on.
Start
The command line, a session and its keys, signing in and choosing a model.
CI and scripts
llx exec and llx decide in a pipeline: the GitHub Action, shell recipes and ready-made workflows.
- llx in CI and scriptsIn CI llx is the same agent as at your terminal: llx exec, with read, edit and bash. It reaches GitHub, GitLab or anything else the way you would, with gh, git, jq and your project's own commands in…
- Recipesllx has no GitHub, GitLab or review feature, and needs none: llx decide and llx exec read stdin and write stdout, so the CLIs you already have read the input and post the answer. Each recipe is one…
- ExamplesGitHub workflows that run llx through its action, rsaryev/llx@v0. Each was run for real in a test repository before it was published.
Reference
The seven tools, modes and approvals, the MCP server, the exit codes and the JSON events.
- Toolsllx has seven tools. The set is deliberately minimal: a new capability goes into one of them (as an action, or a kind of window tab), or is left to the shell.
- Modes, approvals and trustWhat llx asks before it acts, how to stop it, and what it does and does not protect.
- llx as an MCP serverllx mcp serves llx's window to another agent (any client that speaks MCP over stdio): native browser pages and terminal sessions as tabs of one window you watch, grouped by project, with the same…
- ReferenceExit codes, the --json events and llx decide. Both the exit codes and the events are a public contract.