uno-overlays-api
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8) because it instructs the agent to fetch and process external content that defines its operational parameters.
- Ingestion points: According to
SKILL.md, the agent is directed to fetch app metadata, Markdown documentation, and JSON command schemas fromapp.overlays.unoendpoints (e.g.,/apiv2/controlapps/{{token}},/api/info,/api/json, and/model). - Boundary markers: The instructions lack delimiters or 'ignore embedded instructions' warnings for the agent to apply when interpreting responses from these external endpoints.
- Capability inventory: The skill provides the agent with the capability to perform state-changing HTTP PUT requests to the API, which could be leveraged if the agent follows malicious instructions found within the fetched data.
- Sanitization: No validation or sanitization of the content returned by the API is prescribed before it is integrated into the agent's context.
Audit Metadata