Skillv1.0.0

ClawScan security

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

SuspiciousFeb 25, 2026, 11:10 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill does what it says (send code to a remote analyzer) but its runtime instructions and included script will read arbitrary files and POST their contents to a third-party endpoint without warnings or credential checks, creating a real risk of accidental data exfiltration of secrets or private code.
Guidance: This skill legitimately sends code to a remote analyzer (LogicArt). Before installing or using it, consider: only submit non-sensitive code (remove keys, passwords, and private data first); verify the destination (https://logic.art) and its privacy policy; test with small, harmless samples first; if you must analyze private repos, prefer a local/static analyzer or a service that supports private/authenticated uploads; monitor network activity if you want to ensure no unexpected data is being sent. If you need higher assurance, ask the publisher how they handle, store, and delete submitted code and whether they support on-prem or authenticated endpoints.

Purpose & Capability: okName/description (code review via LogicArt) match the included script and SKILL.md: the script reads code or files and POSTs them to https://logic.art/api/agent/analyze for analysis. Requiring file reads and network calls is coherent with a remote code-analysis service.
Instruction Scope: concernSKILL.md and scripts instruct invoking the bundled script or curling the LogicArt API; the analyze.mjs will read any file path passed with --file and send its full contents to an external service. There is no warning about sensitive data or guidance to redact secrets before sending, so normal use could leak credentials or proprietary code.
Install Mechanism: okNo install spec; the skill is instruction-only with a single Node script. Nothing is downloaded from arbitrary URLs and no archive extraction occurs. Low installation risk.
Credentials: noteThe skill requests no environment variables or credentials (which is consistent), but it performs unauthenticated POSTs of user code to a third-party endpoint. Lack of required credentials means the service may accept unauthenticated uploads — useful but increases risk of sensitive-data transmission without access controls.
Persistence & Privilege: okalways is false and there are no instructions to modify other skills, system-wide agent config, or persist credentials. The skill does not request elevated/persistent privileges.