Back to skill
Skillv1.0.0
ClawScan security
clauwdit · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 7, 2026, 5:11 PM
- Verdict
- benign
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill is internally coherent for a remote SKILL.md auditor — it contains only usage instructions that call a third‑party HTTP service and does not request extra credentials or install code — but it relies entirely on an external endpoint, so trust in that site is required.
- Guidance
- This skill is a thin wrapper describing a remote static-audit service — it does what it says, but it relies on sending SKILL.md content to clauwdit.4worlds.dev. Before using: (1) Do not submit any secrets, API keys, or private files; review and redact sensitive data from SKILL.md before posting. (2) Verify the ownership and TLS certificate of clauwdit.4worlds.dev, check for a privacy policy or contact info (the SKILL.md says 'Built by 4Worlds' but the registry shows no homepage). (3) If you need offline analysis or cannot trust an external service, prefer a local/open-source scanner instead. (4) Consider not enabling autonomous invocation for this skill unless you trust the endpoint and understand what data it will receive. If you want higher assurance, ask the publisher for the scanner source code or an official homepage before installing.
Review Dimensions
- Purpose & Capability
- okThe name and description (security auditor for SKILL.md files) match the content: the SKILL.md documents a remote static analysis service and shows how to POST skill contents for analysis. There are no unrelated env vars, binaries, or install steps requested.
- Instruction Scope
- noteThe instructions direct the user/agent to send SKILL.md content to https://clauwdit.4worlds.dev via curl/POST. This is consistent with an online auditor, but it means sensitive skill contents (including any secrets present in a SKILL.md) would be transmitted to a third party. The SKILL.md does not instruct the agent to read other local files or environment variables beyond the content to be submitted.
- Install Mechanism
- okNo install spec or code is included (instruction-only). That minimizes on-disk risk; there are no downloads, extracts, or package installs to review.
- Credentials
- okThe skill does not request any environment variables, credentials, or config paths. There is no disproportionate credential access declared. However, the runtime behavior depends on network access to an external domain.
- Persistence & Privilege
- okalways is false and there is no mechanism that would persist or modify other skills or agent-wide configuration. The skill does not request elevated persistence or unusual privileges.