claude-queue CLAUDE.md

A task queue that lets users group coding-agent tasks, set priorities and dependencies, and run them in the background. It retries tasks and monitors usage limits.

In plain words
What is it for?
Use it to run batches of Claude Code tasks without watching them continuously.
Why use it?
It removes the need to start and supervise each task manually, especially when tasks fail or usage limits interrupt work.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/vasiliyk/claude-queue/claude-md
Clone the repo
git clone --depth 1 https://github.com/vasiliyk/claude-queue
Per session 981 This file is loaded in full into every session.
When invoked 981 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00981 $0.00981
Opus 5 $0.00491 $0.00491
Sonnet 5 $0.00196 $0.00196
Haiku 4.5 $0.00098 $0.00098

Measured 3d ago against content hash 2a9e1257e468, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

claude-queue CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 96 lines

How it starts

The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

What This Is

A task queue system for batching Claude Code tasks with automatic retry and rate-limit handling. Users queue tasks with priorities and dependencies, then run them unattended while the worker monitors Claude Plan usage limits (5-hour session and 7-day weekly quotas).

Execution Model

This is a clone-and-run script, not an installable package. It is invoked either by the user or by Claude Code directly via Bash — never installed as a system command.

Two usage paths:

1. Via the queue-task skill — the skill has allowed-tools: Write, Read only. It cannot run Bash. Its job is to write tasks.yaml and provide a one-liner the user (or Claude Code main agent) can execute:

queue-task skill → writes tasks.yaml → Claude Code / user runs ./claude-queue.py

2. Directly — Claude Code (main agent with Bash access) or the user runs ./claude-queue.py commands directly from the repo root.

pip install -e ".[dev]" is only used to get pytest/ruff onto PATH for development. There is no package to install.

Commands

# Install all dependencies (creates .venv, installs from uv.lock)
uv sync --extra dev

# Run all tests
uv run pytest tests/ -v

# Run a single test file
uv run pytest tests/test_queue.py -v

# Run a single test
uv run pytest tests/test_queue.py::TestTaskQueue::test_add_task -v

# Run with coverage
uv run pytest tests/ --cov --cov-report=html

# Type check
uv run mypy claude-queue.py

# Lint
uv run ruff check .

# Format
uv run ruff format .

# Update lockfile after changing dependencies
uv lock

# Install pre-commit hooks (one-time setup)
uv run pre-commit install

# Run pre-commit manually on all files
uv run pre-commit run --all-files

Architecture

Everything lives in a single monolithic file: claude-queue.py (1384 lines). This is intentional — the tool is cloned and run directly, so a single file is simpler to use and share.

Read the full file on GitHub · 96 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 96 lines · 981 tokens per session scan A 2a9e1257e468

Subscribe to this mod's changes

claude-queue CLAUDE.md is an instructions file published in the GitHub repository vasiliyk/claude-queue (23 stars, last pushed 4mo ago), licensed MIT. It adds 981 tokens to every session, about $0.0049 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.