teach
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates serving lesson content via a local HTTP server using
python3 -m http.server. It also suggests the creation of anopen-lessonscript to automate server startup and browser opening. These operations are restricted to the local development environment. - [EXTERNAL_DOWNLOADS]: The agent uses
WebSearchto find instructional content andcurlto query the YouTube oEmbed API for metadata verification. These are interactions with well-known educational and technology services used to populate lesson resources. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by integrating external web content into local lesson files.
- Ingestion points: WebSearch results and video metadata from external platforms.
- Boundary markers: No specific delimiters or safety warnings are instructed for use when interpolating external content into HTML lessons.
- Capability inventory: Capability includes writing files to the local filesystem and executing local shell commands for server management.
- Sanitization: No explicit sanitization or validation of untrusted external content is required before inclusion in generated materials.
Audit Metadata