Skip to content

Galley

v0.3.0 Feature

This release adds 2 notable features for engineering teams evaluating rollout.

Published 24d AI Coding Tools
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

ai-agents automation claude-code codex daemon developer-tools
+6 more
git-worktree local-first openai orchestration task-runner workflow

Summary

AI summary

Added idle-output watchdog and forced daemon stop command to improve reliability.

Full changelog

🚢 Galley v0.3.0 improves daemon reliability and adds optional acceptance-test skeleton support for AFK tasks.

Highlights

  • Added an idle-output watchdog for executor and built-in supervisor runs, so stalled subprocesses are terminated and retried according to the task loop budget.
  • Added galley daemon stop --force to escalate from graceful shutdown to verified SIGKILL after the stop timeout.
  • Improved daemon restart recovery by requeueing running tasks whose recorded daemon owner is no longer alive.
  • Added optional preflight.acceptance_skeleton support to create AC-linked test skeletons before executor work starts, pass the obligations into the work order, and expose runtime skeleton evidence in task output.
  • Improved restart handling for context-only input files in reused task worktrees.
  • Updated the packaged Claude and Codex Galley plugins to 0.1.4, including task-authoring guidance for AC test skeleton preflight.

Notes

  • Acceptance skeleton preflight is opt-in through task YAML.
  • The required-check acceptance gate now treats preferred_commands as an ordered fallback list, matching the result completion behavior.

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track Galley

Get notified when new releases ship.

Sign up free

About Galley

Local-first runtime for supervised AI coding tasks: isolated git worktrees, supervisor review against acceptance criteria, retry/escalate loops, on-disk run evidence, and PR handoff. Supports Codex CLI and Claude Code. Go, MIT.

All releases →

Beta — feedback welcome: [email protected]