ramp

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses a local configuration file at ~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md. This access is restricted to the skill's own operational data (clinic profile, jurisdiction, and handbook paths) and does not involve harvesting sensitive system credentials or performing network exfiltration.- [INDIRECT_PROMPT_INJECTION]: The skill ingests an external handbook file provided by the professor. While this creates a potential surface for indirect prompt injection, the skill lacks the dangerous capabilities (such as shell execution, network access, or file-writing) necessary for an attacker to achieve high-impact exploitation. The skill also explicitly prompts the student to verify all AI-generated citations and documents.
  • Ingestion points: Reads clinic handbook and configuration from ~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md.
  • Boundary markers: Absent; the skill relies on instructional context to separate data from commands.
  • Capability inventory: No subprocess calls, network operations, or file-write capabilities were identified in the skill instructions.
  • Sanitization: Not present; the skill mitigates risk through explicit student verification requirements and a lack of executable capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 03:58 PM
Security Audit — agent-trust-hub — ramp