contentful-help

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local bash script (scripts/run) to execute a Node.js entry point. This is the standard operational method for this skill's architecture to provide dynamic workflow prompts.
  • [PROMPT_INJECTION]: The skill employs XML tags such as <system> and <prompt> to manage the agent's persona and tasks throughout the workflow. These are architectural components of the skill's engine rather than attempts to bypass security guidelines.
  • [DATA_EXFILTRATION]: While the skill interacts with Contentful environment variables (e.g., CONTENTFUL_SPACE_ID) for configuration, no patterns of unauthorized data harvesting or network exfiltration were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 08:29 PM
Security Audit — agent-trust-hub — contentful-help