gsap-core
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied source material and external data, which creates a potential surface for indirect prompt injection.
- Ingestion points: Processes user-provided stack, source material, and inputs as defined in SKILL.md.
- Boundary markers: SKILL.md instructs the agent to preserve authorization boundaries, though no specific text delimiters for untrusted input are mandated.
- Capability inventory: The skill utilizes sandbase tools in SKILL.md and sandbase-api-map.md for external media generation.
- Sanitization: SKILL.md includes a robust mitigation requiring the agent to present all parameters and costs to the user for explicit confirmation before any external call is executed.
Audit Metadata