memo

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a template generator for legal education and does not possess capabilities for network communication, arbitrary code execution, or unauthorized file access.
  • [DATA_EXPOSURE]: The skill references configuration paths under ~/.claude/plugins/config/claude-for-legal/. These are localized configuration settings for the skill's specific environment and do not expose sensitive system or user credentials.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of case notes and summaries. It mitigates potential risks by using clear output boundaries (bracketed placeholders like [学生分析]) and by restricting its capabilities solely to text generation, lacking the ability to perform network or system-level actions.
  • [INDIRECT_PROMPT_INJECTION]: Mandatory Evidence Chain: 1. Ingestion points: Reception summaries and case notes (SKILL.md). 2. Boundary markers: Explicit bracketed markers for AI prompts and student placeholders. 3. Capability inventory: Text generation only; no subprocess, network, or file-write tools are invoked. 4. Sanitization: The skill mandates that users verify all citations against primary legal sources before use.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 01:02 AM
Security Audit — agent-trust-hub — memo