You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This commit was created on GitHub.com and signed with GitHub’s verified signature.
Fixed
No-spend proof runs now complete honestly - martin-loop run ... --proof no longer misclassifies a clean verifier-only pass as no_code_change, which previously cascaded into a false budget_exit.
Portable Windows Codex execution - Windows command resolution now handles npm-style .cmd shims correctly, including real codex.cmd and verifier commands launched through npm wrappers.
Hosted OpenAI defaults - --engine openai now targets the hosted OpenAI endpoint by default, so local proofs do not require MARTIN_OPENAI_BASE_URL just to avoid a localhost misroute.
Claude live runs honor provider permissions again - the public Claude adapter no longer force-adds --dangerously-skip-permissions on live governed runs.
Changed
Public CLI guidance now consistently uses martin-loop in help, tour, onboarding, workflow receipts, and MCP config hints.
The public root tarball no longer exposes the vendored internal CLI bin surface inside dist/vendor/cli/package.json.
