compress

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted input to perform compression, creating a surface for indirect prompt injection attacks.
  • Ingestion points: Untrusted data enters the agent context through $ARGUMENTS, pasted text, or file content specified in SKILL.md.
  • Boundary markers: Absent. The prompt does not define delimiters or provide specific instructions to the agent to disregard commands embedded within the text being compressed.
  • Capability inventory: The skill instructions (Step 5) describe replacing file contents, which implies the agent has file-write permissions.
  • Sanitization: Absent. The skill does not specify any validation, escaping, or filtering of the input before processing it.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 10:52 AM
Security Audit — agent-trust-hub — compress