desiderio-teach

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell utilities find and rg (ripgrep) to search for configuration files (config.yaml) and templates (*.fluid.html) within the project's ContentBlocks/ContentElements directory. These commands are used solely for local introspection to map content elements correctly.
  • [EXTERNAL_DOWNLOADS]: The skill references documentation and attribution from a well-known GitHub repository (mattpocock/skills) and the author's official domain (webconsulting.at). These are standard references for developer tools and do not involve untrusted remote code execution.
  • [PROMPT_INJECTION]: The skill processes external information through the creation of RESOURCES.md and by reading local project files. While this presents a surface for indirect prompt injection, the skill mitigates this by instructing the agent to ground all teaching claims in primary documentation and local project context, providing a verification checklist for all generated content.
  • [DATA_EXFILTRATION]: No patterns for exfiltration of sensitive data were detected. The skill focuses on local project file creation (MISSION.md, RESOURCES.md, NOTES.md) and standard TYPO3 CMS record management through official APIs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 07:25 AM
Security Audit — agent-trust-hub — desiderio-teach