agent-security-hardening

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a set of security guidelines and defensive programming examples designed to help developers build secure AI agent frameworks. It does not contain any executable malicious code.
  • [COMMAND_EXECUTION]: The skill includes defensive code for validating agent identifiers used in CLI arguments, advising the use of end-of-options markers and argv APIs to prevent argument injection.
  • [DATA_EXFILTRATION]: Guidelines are provided for protecting credentials and sensitive data, including managing environment variables securely and ensuring secrets are not leaked into logs or command histories.
  • [PROMPT_INJECTION]: The document explicitly addresses prompt injection by advising developers to move security invariants from prompts into deterministic code, treating model outputs as untrusted input.
  • [DYNAMIC_EXECUTION]: The skill provides robust path containment logic using Python's pathlib to resolve workspace boundaries and prevent directory traversal attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 12:25 AM