pentest-input-protocol-manipulation

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE] (SAFE): The skill consists entirely of textual instructions for the agent to follow during a security assessment. No malicious logic, prompt injection attempts, or unauthorized behaviors were detected.
  • [NO_CODE] (SAFE): No scripts, binaries, or package manager files (e.g., package.json or requirements.txt) are included, eliminating risks associated with code execution or remote downloads.
  • [CREDENTIALS_UNSAFE] (SAFE): The file does not contain any hardcoded secrets, API keys, or sensitive environmental paths.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 03:06 PM