Build and operate a multi-agent task queue in OpenClaw with priority queues, delayed/dead-letter queues, scheduling, retry/timeout control, dependency manage...
Security Analysis
high confidenceThe package implements a local TypeScript multi-agent task queue whose code, instructions, and dependency requests are coherent with the stated purpose.
The name/description (multi-agent task queue) matches the included TypeScript runtime (TaskQueue, Scheduler, DependencyManager, and storage backends). The package.json dependencies (redis, better-sqlite3) align with the provided Redis and SQLite storage implementations.
SKILL.md instructs only queue-related actions (importing TaskQueue/Scheduler, picking storage backend, registering handlers, enqueueing tasks, running scheduler). It advises running npm install / npm run check; it does not instruct reading unrelated files, environment variables, or posting data to unknown endpoints.
There is no automated install spec in the registry bundle; the README/SKILL.md expect you to run npm install locally. Dependencies are from npm and there are no downloads from arbitrary URLs. Note: better-sqlite3 is a native module that may require build tools when installed.
The skill declares no required env vars or credentials. Redis usage is opt-in via the RedisStorage constructor (accepts a URL argument) — no hidden env access is present. No unrelated secrets are requested.
The skill is not forced-always, does not request elevated platform privileges, and does not modify other skills or global agent configuration. Running the SQLite backend will create a local DB file; using Redis will connect to a provided Redis endpoint.
Guidance
This skill appears to be what it claims: a bundled TypeScript task-queue runtime. Before installing or running it: (1) inspect package.json and be prepared to run npm install (better-sqlite3 is a native module and may require build tools); (2) if you use RedisStorage, only point it at Redis instances you trust because task data and dependency results will be stored there; (3) SQLiteStorage will create a local file (task-queue.sqlite) in the working directory; (4) running npm test or npm run check will execute the test suite and build steps — run these in a sandbox if you are cautious; (5) note that task handlers you register execute arbitrary code and therefore can access network, files, or secrets available to the runtime; review any handlers you or the agent provides before allowing the scheduler to run continuously.
Latest Release
v0.1.0
Initial release
More by @imgolye
Published by @imgolye on ClawHub
