security-best-practices

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains no executable scripts, binaries, or automated installers. It is composed of markdown-based security specifications that instruct the AI agent on how to securely review user code.
  • [DATA_EXFILTRATION]: All reference files (e.g., golang-general-backend-security.md, javascript-express-web-server-security.md) include explicit 'MUST NOT' requirements regarding secrets. The agent is strictly forbidden from requesting, logging, or committing API keys, passwords, private keys, or session tokens.
  • [PROMPT_INJECTION]: The skill includes instructions for handling 'Overrides,' allowing the agent to bypass certain best practices if the project context requires it. This is evaluated as legitimate functional guidance for a coding assistant to handle legacy or specialized environments and does not target AI safety filters or attempt to subvert the agent's core safety guidelines.
  • [REMOTE_CODE_EXECUTION]: No remote code execution or untrusted external download patterns were detected. All external links point to trusted, well-known technical and security documentation sources (e.g., OWASP, MDN, and official framework repositories).
  • [EXTERNAL_DOWNLOADS]: The skill mentions external tools like govulncheck and various NPM/PyPI libraries (e.g., helmet, zod, bcrypt) in an advisory capacity as recommendations for the user's codebase, rather than as dependencies to be installed by the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 08:24 PM