worked-example-walkthrough
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents a potential for indirect prompt injection because it processes untrusted user data with accompanying execution capabilities.
- Ingestion points: The skill ingests user-specified mathematical problems and parameters to generate its walkthrough frames, as seen in
SKILL.md. - Boundary markers: No specific boundary markers or instructions to ignore embedded prompts in user data are provided.
- Capability inventory: The 'Guardrails' section in
SKILL.mdexplicitly instructs the agent to use aBashshell andnumpyto compute states for complex problems. - Sanitization: No input validation or sanitization routines are specified for the user-provided math problems before they reach the computational stage.
- [COMMAND_EXECUTION]: The skill encourages using a
Bashshell to executenumpyscripts for verifying math operations. Although intended for accuracy, this provides a pathway for code execution derived from untrusted inputs.
Audit Metadata