teach
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to "open the lesson file for the user by running a CLI command." This grants the agent shell command execution capabilities, intended for utility actions like opening HTML documents in the user's environment.
- [EXTERNAL_DOWNLOADS]: The agent is tasked with finding and reading "high-quality resources" from the web to acquire knowledge. This involves network activity and the ingestion of content from untrusted third-party domains.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it processes untrusted data from external websites to generate local files (lessons, records, and glossaries). The instructions do not specify boundary markers or sanitization for this external content, which could potentially influence the agent's behavior during the file creation process.
Audit Metadata