Files
beads/commands/stale.md
Steve Yegge 7658c4a8e8 Add command documentation and improve daemon UX
- Added human-readable uptime formatting (51m 59s vs 3119.4 seconds)
- Fixed daemon stop race condition with SIGKILL
- Added markdown docs for 19 commands: blocked, comments, compact, daemon,
  delete, dep, epic, export, import, label, list, quickstart, rename-prefix,
  renumber, reopen, repos, restore, stale, sync
- Closed: bd-159, bd-152, bd-168
- Deleted test issues: bd-144, bd-145

Amp-Thread-ID: https://ampcode.com/threads/T-9f7c3fed-62de-4bcd-a059-8c1b77cdb841
Co-authored-by: Amp <amp@ampcode.com>
2025-10-19 22:24:14 -07:00

26 lines
761 B
Markdown

---
description: Show orphaned claims and dead executors
argument-hint: [--release] [--threshold]
---
Show issues stuck in_progress with execution_state where the executor is dead or stopped.
Helps identify orphaned work that needs manual recovery.
## Stale Detection
An issue is stale if:
- It has an execution_state (claimed by an executor)
- AND the executor status is 'stopped'
- OR the executor's last_heartbeat is older than threshold
Default threshold: 300 seconds (5 minutes)
## Usage
- **List stale issues**: `bd stale`
- **Custom threshold**: `bd stale --threshold 600` (10 minutes)
- **Auto-release**: `bd stale --release` (automatically release all stale issues)
Useful for parallel execution systems where workers may crash or get stopped.