queue-status
Skill0 starsChanged 5 days agoMITView on GitHub
Reports queue counts and flags missing gates, handles and broken deps. Use when the user asks how the queue stands. Not for changing task status.
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- claude-arsenal/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- coverage-gapsSkill
- dep-upgradeSkill
- designSkill
- executionSkill
- explore-ideaSkill
- gate-checkSkill
- githubSkill
- initSkill
- litcriticSkill
- mutmut-reportSkill
- pin-checkSkill
- pypi-releaseSkill
- python-bootstrapSkill
- queue-addSkill
- queue-nextSkill
- reviewSkill
- session-endSkill
- shipSkill
- specifySkill
The file
Files
queue-status
Reports the board: how many tasks are open, claimed, done and blocked, and —
with --detail — every task with its priority, state, and what holds it up.
CANARY: queue-status-loaded-2026-06-13-fb78d23e-d4e5f6a7b8c9d0e1
When to load
- The user asks how the queue stands, what tasks are left, or types
/queue-status. - Checking whether every task is done before closing a loop session.
- Diagnosing a stuck queue.
How to use
Fetch the arsenal:task issues, open and closed, with whatever GitHub access
this surface has, asking for number, title, state, labels, assignees
(plus state_reason where the surface returns it) and leaving out body — the board is derived from labels and state, and bodies
are most of the fetch. Then run query_status.py (in claude-arsenal/scripts/):
query_status.py --issues /tmp/issues.json # summary counts
query_status.py --issues /tmp/issues.json --detail # every task with blockersIt reads the same graph as the selector, so the board matches what a worker
will pick next. Without --issues every task shows as open, because state
lives in the issues.
What it flags
Problems go to stderr; --fail-on-problems makes them a non-zero exit for a
make target or CI job.
no-gate— no fenced```bashblock, so the gate runs nothing and passes everything.no-handle— no issue points at the task, so no agent can claim it. Runhandle_sync.pyand create the missing issues; where it reports a near-identical issue title, addarsenal-task: <id>to that issue instead of opening a second.depends on unknown task— a dep id no task file declares. The selector treats it as unsatisfied, so the task would never become eligible.- completion drift — the task file and its issue disagree on whether the work is finished (a merged file with an open issue, or a completed issue with a live file). Selection hides this on purpose, so the board is where it shows.
Reading the board
blockedis not failed: the task becomes eligible once its dependencies are closed as completed. A close as not-planned leaves dependents blocked.claimedmeans an agent holds the claim ref; the issue's assignee and claim comment name the session. A claim over a day old with no open PR may be stale;arsenal-queue.ymlreleases those on a schedule, andclaude-arsenal:core:queue-next § references/claim-gotchas.mdsays how stale claims are released.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

