teach
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mdsuggest that the agent should run CLI commands to open generated HTML course files for the user. - [EXTERNAL_DOWNLOADS]: The agent is directed to fetch and utilize knowledge from external sources and high-trust resources identified in
RESOURCES.mdto generate content, rather than relying solely on its internal knowledge. - [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection through the ingestion of untrusted data.
- Ingestion points: Data enters the context via external URLs and descriptions provided in
RESOURCES.md, as well as user-defined content inMISSION.mdandNOTES.md. - Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when processing external content.
- Capability inventory: The skill allows the agent to write files to the local workspace (
./lessons/*.html,./learning-records/*.md) and execute shell commands to open files. - Sanitization: No sanitization or validation of external content is mentioned before it is interpolated into lessons or stored in the workspace.
Audit Metadata