eli5
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided topics to generate HTML explainers. While this involves processing untrusted data, the risk is minimized by the skill's specific formatting requirements.
- Ingestion points: The skill ingests a topic from the user's request (SKILL.md).
- Boundary markers: Explicit rules for ELI5 simplification and visual structure limit the potential for the model to follow instructions embedded within the topic.
- Capability inventory: The skill can save HTML files or return code blocks, but it explicitly forbids network access or external package dependencies for the output.
- Sanitization: The skill enforces a "no network access" policy for the generated HTML, preventing the use of external scripts or tracking pixels.
- [EXTERNAL_DOWNLOADS]: The README references the author's own domain and GitHub repository for installation, as well as an upstream repository from a trusted organization for attribution. These are standard distribution practices and do not involve unauthorized code execution.
Audit Metadata