teach
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests data from external sources and community forums to generate educational content. This creates a surface for indirect prompt injection where untrusted data could influence agent behavior.
- Ingestion points: Data is pulled from
RESOURCES.md, external URLs, and community forums. - Boundary markers: Absent; the instructions do not specify the use of delimiters or warnings to ignore embedded instructions within processed data.
- Capability inventory: The skill performs file writes to multiple directories (
lessons/,reference/,assets/,learning-records/) and is explicitly prompted to execute CLI commands. - Sanitization: There is no mention of sanitizing, escaping, or validating external content before it is used in lessons or decision-making.
- [COMMAND_EXECUTION]: The instructions explicitly tell the agent: 'If possible, open the lesson file for the user by running a CLI command.' While the intended purpose is the legitimate opening of generated HTML files, this constitutes an instruction to interact directly with the shell environment, which could be exploited to run unauthorized commands if the path or command is not strictly controlled.
Audit Metadata