apply-anthropic-skill-best-practices

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a static educational resource and does not contain any executable logic or instructions that bypass security controls.
  • [EXTERNAL_DOWNLOADS]: The content references official documentation from docs.claude.com and documentation assets on mintcdn.com. These are legitimate and appropriate sources for the skill's instructional purpose.
  • [COMMAND_EXECUTION]: The document contains example shell commands (e.g., grep, pip install, python scripts/...) used strictly to illustrate development workflows. These are descriptive and do not initiate unauthorized operations.
  • [PYTHON_PACKAGES]: The examples mention standard open-source libraries such as pdfplumber, pypdf, pytesseract, and pdf2image, which are typical for the data processing tasks discussed.
  • [REMOTE_CODE_EXECUTION]: The skill mentions the platform's ability to install packages from npm and PyPI for informational purposes; no malicious or unverified remote execution patterns were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:10 AM
Security Audit — agent-trust-hub — apply-anthropic-skill-best-practices