Skillv1.0.1

ClawScan security

Skill Update Helper · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 12, 2026, 11:11 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's instructions, lack of installs, and requested resources match its stated purpose of checking and applying OpenClaw/skill updates; nothing requests unrelated credentials or odd installs.
Guidance: This skill appears coherent and focused on update workflows. Before using or scheduling it: (1) confirm you have clawhub/openclaw installed and that you want automatic updates; (2) prefer dry-run (clawhub update --all --dry-run) first; (3) avoid blindly running 'clawhub update --all' in production—consider updating specific skills or testing in staging; (4) if you enable a cron job, explicitly approve its schedule and command and ensure you have backups/rollback procedures; (5) verify any credential prompts come from your existing clawhub/OpenClaw tooling (use 'clawhub whoami' to check).

Purpose & Capability: okName/description match the instructions: all commands and workflow reference OpenClaw tooling (clawhub, openclaw) and scheduling updates, which is appropriate for an update helper. No unrelated binaries, env vars, or config paths are required.
Instruction Scope: okSKILL.md and agent-guide instruct only operations relevant to update checks and scheduling (openclaw status, clawhub list/update, use cron for scheduling). They do not direct reading unrelated files, exfiltrating data, or calling external endpoints outside OpenClaw tooling. The guidance to check auth (clawhub whoami) is reasonable.
Install Mechanism: okInstruction-only skill with no install spec and no code files — minimal footprint and nothing is written to disk by the skill itself.
Credentials: okNo environment variables, credentials, or config paths are requested. The operational commands may require the user's existing clawhub/OpenClaw credentials, but the skill does not demand unrelated secrets.
Persistence & Privilege: noteThe skill recommends creating scheduled cron tasks to run automatic update checks; while the skill itself is not forcibly persistent (always:false), scheduling introduces persistent, autonomous behavior. This is expected for an update helper but worth explicit user consent.