understand-explain

Pass

Audited by Gen Agent Trust Hub on Apr 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface by interpolating user-provided arguments ($ARGUMENTS) directly into the agent's search instructions. While designed for looking up file paths and function names, this pattern allows untrusted data to influence the agent's logic.
  • Ingestion points: User-provided component identifier via $ARGUMENTS (SKILL.md)
  • Boundary markers: Absent; the argument is placed directly into the search context.
  • Capability inventory: Mentions 'Grep' and file reading capabilities to access the knowledge graph and source code (SKILL.md).
  • Sanitization: Absent; the skill does not instruct the agent to validate or sanitize the input before using it in search commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 18, 2026, 12:07 PM
Security Audit — agent-trust-hub — understand-explain