Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Agent Os
v0.1.0Persistent agent operating system for OpenClaw. Agents remember across sessions, learn from experience, coordinate on complex projects without duplicate work.
⭐ 0· 547·1 current·1 all-time
byNovaiok@cryptocana
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description describe a persistent multi-agent orchestrator; the repository contains Agent, TaskRouter, Executor, index, UI and example files that implement exactly that functionality. There are no unrelated required binaries or environment variables.
Instruction Scope
Runtime instructions and API examples are consistent with the code. The SKILL.md and code explicitly state that state is persisted to a local data/ directory and the code reads/writes JSON files there; this is expected for a persistent agent OS but you should be aware project data (goals, outputs, agent memories) will be written to disk under the skill's directory.
Install Mechanism
This is an instruction-and-code package with no install spec in the skill metadata. No remote installers, downloads, or external release hosts are used in the metadata provided. The presence of package.json implies npm-style usage; review package.json for scripts before running installs.
Credentials
No environment variables, credentials, or config paths are required. The code only uses filesystem I/O under the skill's ../data path and standard Node.js modules (fs, path). There are no requests for unrelated tokens or secrets.
Persistence & Privilege
The skill persists state to disk (data/*.json) which is necessary for its stated goal. It does not request 'always: true' and uses default autonomous-invocation settings. Autonomous invocation is normal; because this skill stores persistent project and agent data locally, consider whether you want that data persisted on the host running the agent.
Assessment
This skill appears coherent: it implements a local persistent multi-agent orchestrator and only writes JSON state files under its data/ directory. Before installing, review package.json (for any postinstall scripts), and confirm you are comfortable with project/agent data being stored on the host where the skill runs. If you plan to run real agents or connect to external AI services later, double-check any future changes that introduce network calls or credential requirements. If you need stricter isolation, run the skill in a sandboxed environment or review/modify the paths where it writes data.Like a lobster shell, security has layers — review code before you run it.
latestvk977bxcp9htpet90dsdwbyzyv981rsmm
License
MIT-0
Free to use, modify, and redistribute. No attribution required.