Skip to content

claudeloop

Onion-architected, autonomous Claude Code session runner and full Anthropic SDK CLI — never blocks on a human, distinguishes an exhausted rate-limit window from exhausted credits, and resumes safely across usage windows.

What problem this solves

Claude Code sessions hit usage limits. A claude -p invocation ending doesn't tell you whether the task finished or just that turn did. And when a rate limit rejects you, you can't tell from the outside whether waiting will ever help — a five-hour window will reset on its own; an exhausted credits balance never will, no matter how long you wait.

claudeloop exists to get all three of those distinctions right, automatically, so you can hand it a plan and walk away — including handling the case where you top up your account's credits while it's mid-wait.

Where to go next

I want to... Read
Install it and run my first autonomous session Getting started
Understand how rate limits and credits are told apart Rate limits vs. credits
Understand why it can never get stuck waiting on me Never blocking
See the full system design Architecture overview
See why a specific hard call was made Architecture decision records
Contribute code Development guide
See the original plans this project was built from Plans

Project status

Pre-1.0. Milestone M1 (the pure domain core) is complete — a fully tested set of value objects and a pure state machine that model every hard decision the runner makes. Milestones M2–M5 (the actual agent integration, the CLI, the generated REST surface, and final polish) are roadmap. See plans/architecture-and-roadmap.md for the build order.

License

MIT. See LICENSE.