research-start
Pass
Audited by Gen Agent Trust Hub on Jul 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for educational legal research guidance. It facilitates the creation of research roadmaps (laws, case areas, keywords) without providing authoritative citations, requiring manual student verification.
- [SAFE]: File access is restricted to a local configuration file (
~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md) used to identify jurisdiction and practice areas. This is a legitimate functional requirement for a legal clinical education tool. - [DATA_EXFILTRATION]: No network-enabled tools or commands are present. The skill only generates text output for the user, preventing any risk of secret exfiltration.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests external data (user-uploaded research and 'seed files' referenced in configuration), it possesses no dangerous capabilities such as shell access, file writing, or network connectivity. Therefore, the risk of cross-context instruction execution is negligible.
- [SAFE]: The skill explicitly includes defensive instructions against AI hallucinations, specifically prohibiting 'silent supplementation' and requiring clear labeling of the source of every citation (e.g., [Model Knowledge], [User Provided]).
Audit Metadata