feature-risk-assessment

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • [Contextual File Access]: The skill accesses local configuration and matter files (e.g., CLAUDE.md, matter.md) stored within a specific directory structure (~/.claude/plugins/config/claude-for-legal/). This is a standard operation within this environment to provide relevant, matter-specific legal context.
  • [Mitigation of AI Hallucinations]: The instructions include a robust 'Citation check' section that explicitly warns the user about the risk of AI-fabricated legal citations. It mandates verification against primary sources like Westlaw or CourtListener before final use, which is a security best practice for AI-assisted legal work.
  • [Data Segregation]: The skill includes logic to prevent 'cross-matter context' contamination by ensuring that files from one legal matter are not read when working on another unless explicitly permitted.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 12:40 AM