teach
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and explain untrusted data including source code, git diffs, and outputs from other automated skills.
- Ingestion points: Instructions in
SKILL.mddirect the agent to read source code, git diffs, and the output of thehowandwhytools. - Boundary markers: The instructions do not define specific delimiters or instructions to prevent the agent from obeying instructions embedded within the code or diffs being analyzed.
- Capability inventory: The skill uses a debugger, git tools, and image generation, and it invokes downstream skills, creating a potential path for data access or tool abuse if triggered by malicious input data.
- Sanitization: There is no mention of sanitization or filtering for the external code or diff content before it is processed by the LLM.
Audit Metadata