Headless agent
A headless agent runs without an interactive interface — invoked by a script, a CI job, or a scheduler rather than a person typing. It takes a task as input, works autonomously, and produces output such as a diff, a pull request, or a report.
Headless operation is what turns an agent from a tool into infrastructure. Instead of asking it to fix the failing test, a CI job invokes it whenever a test fails and opens a pull request with a proposed fix.
The requirements are stricter than interactive use. There is nobody to answer a clarifying question, so the prompt must be complete. There is nobody to catch a bad change, so the verification step is doing real work. And the boundary matters more, because an unsupervised agent with a vague scope is how surprising things get committed.
What this means in practice
Claude Code supports non-interactive invocation, which is the usual starting point for this. Begin with tasks that have machine-checkable success conditions — make the suite pass, fix the type errors — before anything requiring judgement.
Prompts for the tools this applies to
Related terms
- CLI coding agent
A CLI coding agent is a coding agent you invoke from the command line.
- Terminal AI agent
A terminal AI agent runs in your shell rather than inside an editor.
- Autonomous coding agent
An autonomous coding agent completes a described task without step-by-step supervision: it plans, edits, runs, and verifies on its own, reporting back when done.