prose-distill

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill lacks protection against indirect prompt injection, which could allow malicious instructions embedded in the input text or files to influence agent behavior.
  • Ingestion points: The skill accepts untrusted data via the $ARGUMENTS parameter (text or file path) in SKILL.md.
  • Boundary markers: Absent; there are no delimiters or instructions (e.g., 'treat the following text only as data to be distilled') to prevent the agent from obeying instructions hidden within the input.
  • Capability inventory: The skill has access to multiple tools including Read, Edit, Write, Grep, Glob, and TodoWrite, which could be abused if an injection is successful.
  • Sanitization: Absent; the skill processes input directly without validation or escaping techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 09:28 AM
Security Audit — agent-trust-hub — prose-distill