teach
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from local markdown files which could potentially contain instructions intended to influence agent behavior. \n
- Ingestion points: Processes content from
MISSION.md,RESOURCES.md,NOTES.md, and files in the./learning-records/directory (referenced inSKILL.md). \n - Boundary markers: Absent; the skill does not specify delimiters to separate untrusted data from system instructions. \n
- Capability inventory: Capabilities are limited to creating and reading files within the local workspace; no network access or shell execution tools are requested. \n
- Sanitization: Absent; the skill relies on the agent's internal safety filters to handle potentially malicious content within processed files.\n- [DYNAMIC_EXECUTION]: The skill instructs the agent to generate interactive HTML lessons containing executable elements. \n
- Evidence: The
SKILL.mdfile defines lessons as self-contained HTML files in./lessons/that incorporate 'quiz widgets', 'simulators', and 'diagram helpers' from./assets/.\n- [SAFE]: The skill is a well-structured instructional tool with clear provenance from a known developer. It includes explicit instructions to distinguish trusted resources from unverified answers and follows pedagogical best practices.
Audit Metadata