v0.1.0 — Initial public release

v0.1.0 — Initial public release

agent-loop helps existing coding agents carry research and engineering projects through repeated planning,
execution, verification, and review. This is an early release for users willing to report workflow and
installation issues.

Included

• Reviewed plans with revision-bound tasks, runtime pause/resume, and Goal Review.
• A bounded command runner, execution evidence, and an independent project verifier.
• Optional plan/task reviewers, project memory, and record retirement policies.
• A Windows Cursor adapter; an optional phone pager distributed as a separate package.
• English and Chinese entry docs, a reproducible failing-to-passing quickstart, and isolated package installation checks.

Get started

Clone the repository, check out v0.1.0, and run python -m pip install .. Follow
Quickstart to verify the CLI without a
model or API key, then Usage to configure a working project.

Support and limitations

Python 3.11+; CI targets Windows and Ubuntu on Python 3.11 / 3.13. The supported host installer is Windows +
PowerShell for Cursor. Other hosts may use portable components, without a guaranteed completion gate.
Only explicitly delegated commands are controlled by the runner. Pager cannot wake an exited agent.
Configuration/state formats may change before 1.0; consult the changelog before upgrading.

The full setup is distributed through the repository tag. The Python package contains CLI/adapter code;
rules, skills, installer scripts, and pager remain in the checkout. No PyPI publication is required for this release.

Feedback

Please report installation/host compatibility problems, incorrect completion decisions, repeated work, and
review overhead through Issues. Include the release/commit,
OS, Python and host versions, and a minimal sanitized reproduction. Use the private security channel for vulnerabilities.