oss-review
Pass
Audited by Gen Agent Trust Hub on Jun 12, 2026
Risk Level: SAFE
Full Analysis
- File System Access: The skill reads local files such as
~/.claude/plugins/config/claude-for-legal/ip-legal/CLAUDE.md, project manifests (package.json,requirements.txt), and source code directories. This access is necessary to evaluate the project's practice profile and license compliance requirements. - Indirect Prompt Injection Surface: The skill is designed to ingest and analyze untrusted external data, specifically package manifests, SBOMs, and license text. This creates a potential surface for indirect prompt injection if those files contain instructions meant to influence the agent's output. However, the skill implements a strict structured workflow, uses a specific output template, and includes requirements for attorney review of sensitive findings, which helps mitigate the risk of unintended obedience to embedded instructions.
- Capability Inventory: The skill's primary capabilities include reading local files and generating structured compliance memos. It does not perform network operations (exfiltration) or execute code derived from the files it reviews.
- Sanitization and Boundaries: The instructions mandate reading the actual license text rather than metadata and require tagging sources with markers like
[verify]. While it does not specify character-level escaping, the high-level process is designed for forensic accuracy and verification.
Audit Metadata