Back to skill
Skillv1.0.11
ClawScan security
Legal Cog · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 23, 2026, 6:19 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's requirements and runtime instructions align with its stated purpose (using the CellCog API to generate legal documents); it only asks for python3 and a CellCog API key and contains no install script or hidden actions.
- Guidance
- This appears coherent and limited: it needs only python3 and your CellCog API key to call the CellCog service. Before installing, confirm the skill's publisher and that you trust https://cellcog.ai (the skill's homepage) because your prompts and generated documents will be sent to that external service. If you want to avoid network calls, do not provide your API key. Also remember AI output is not a substitute for a licensed attorney — review generated legal documents with counsel for important matters.
Review Dimensions
- Purpose & Capability
- okName/description (legal document generation via CellCog) match the declared runtime needs: python3 for the Python SDK snippets and CELLCOG_API_KEY for the external CellCog service. The frontmatter lists a 'cellcog' dependency which is consistent with using the CellCog SDK.
- Instruction Scope
- okSKILL.md contains only instructions and example code to call the CellCog API for drafting documents. It does not instruct the agent to read unrelated files, system config, or additional environment variables beyond CELLCOG_API_KEY, nor to transmit data to unexpected endpoints.
- Install Mechanism
- okNo install spec is present (instruction-only), so nothing is downloaded or written to disk by the skill itself. This minimizes install-time risk.
- Credentials
- okOnly one required env var (CELLCOG_API_KEY) is declared, which is proportionate for a skill that calls an external API. No unrelated credentials or config paths are requested.
- Persistence & Privilege
- okalways is false and the skill does not request elevated persistence or modify other skills. Autonomous invocation is allowed by default (platform behavior) but not combined with other concerning privileges here.