Skillv1.0.0

ClawScan security

Local Healthcheck · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 14, 2026, 8:55 AM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's files and runtime instructions match its stated purpose: it runs local checks (firewall, open ports, updates, SSH) and writes a local report; nothing in the bundle appears to exfiltrate data or require unrelated credentials.
Guidance: This skill appears to do what it says: run local checks and write a report to memory/healthcheck-<date>. Before installing or running: review the script (it is short and readable), be aware it may prompt for sudo to run commands that require root (listing ports, checking apt updates), and note the report is written to a memory/ subdirectory under the run directory. It does not contact the network or request credentials. If you are uncomfortable granting sudo, run the script manually in a controlled environment or remove the sudo calls where appropriate.

Purpose & Capability: okName/description (local security checks) align with the included script and instructions. The script runs local commands to check firewall status, open ports, updates, and SSH — all consistent with the declared purpose. No unexpected services or credentials are requested.
Instruction Scope: okSKILL.md limits behavior to running the checks and writing a report to memory/healthcheck-<date>. The script only reads local system state (commands like firewallctl/socketfilterfw, lsof/ss, apt/softwareupdate, pgrep) and writes a single local report file. It does not contact external endpoints or read unrelated files.
Install Mechanism: okNo install spec. This is an instruction-only skill with a single bundled script (check.sh). Nothing is downloaded or installed from external URLs.
Credentials: okThe skill declares no required env vars or credentials, which matches the script. The script does use sudo for some commands (ufw, ss, lsof, apt) which is appropriate for listing system ports and package updates on many systems; this is a privilege requirement, not a hidden credential request.
Persistence & Privilege: okalways is false and the skill does not attempt to modify agent configuration or other skills. It writes reports to a local memory/ subpath within the current working directory only. The agent-autonomy default is allowed but not unusual; combined with the limited scope this does not raise special persistence concerns.