legal-risk-assessment-anthropic

Warn

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: MEDIUMPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill metadata identifies the author as "Anthropic", which contradicts the provided authorship context indicating the author is "lawve-ai". This misleading attribution is a form of metadata poisoning that may lead users to assign undue trust to the skill's recommendations.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted user data into a structured risk assessment framework without using boundary markers.
  • Ingestion points: Matter descriptions, background context, and risk analysis sections in SKILL.md.
  • Boundary markers: Absent; there are no delimiters or instructions provided to the agent to treat user-provided data as potentially untrusted.
  • Capability inventory: The skill does not define any external tools, file system access, or network capabilities.
  • Sanitization: No input validation or sanitization logic is implemented.
  • [NO_CODE]: The skill consists entirely of markdown documentation and legal frameworks without any executable scripts or tool configurations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 16, 2026, 04:13 AM
Security Audit — agent-trust-hub — legal-risk-assessment-anthropic