dockkit
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references multiple documentation and video resources hosted on the non-trusted domain
sosumi.ai(e.g.,https://sosumi.ai/documentation/dockkit). This domain is not a recognized trusted vendor or well-known service. - [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting untrusted camera frame data to control physical hardware motors, establishing an indirect prompt injection surface.
- Ingestion points: Camera frame data (
CMSampleBuffer) and derivedObservationobjects processed inprocessFrameanddetectAnimalfunctions. - Boundary markers: None. The instructions do not provide delimiters or logic to ignore potential commands embedded in the visual stream.
- Capability inventory: Physical motor control capabilities including
setAngularVelocity,setOrientation, andanimatecalls which drive device movement. - Sanitization: None. The skill lacks validation or filtering of the visual camera data before it influences device behavior.
Audit Metadata