sdk-install
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a coordinator for SDK installation and initialization. It emphasizes security by instructing the agent never to hardcode SDK keys, mobile keys, or client-side IDs, instead directing the use of environment variables and secret management.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a workflow that ingests project context to detect stack information. 1. Ingestion points: repository stack detection in the detect sub-skill. 2. Boundary markers: the workflow includes explicit blocking decision points (D) requiring user confirmation before proceeding. 3. Capability inventory: file system modifications and package installations are delegated to the apply sub-skill. 4. Sanitization: the skill uses pre-defined SDK recipes and snippets to ensure generated code follows established safe patterns.
Audit Metadata