iblai-api-external-service-proxy
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill facilitates interaction with the ibl.ai External Service Proxy. All network requests are directed to the vendor's official API domain (api.iblai.app) or reputable third-party AI providers such as ElevenLabs and HeyGen.
- [SAFE]: Sensitive provider credentials are managed server-side at the organizational level. The skill instructions and reference materials explicitly state that provider API keys are never held by the client, which significantly reduces the risk of credential exposure.
- [SAFE]: Safety protocols are integrated into the instructions, specifically requiring the agent to obtain user confirmation before executing any billable actions (like generating speech or video) or destructive operations (like deleting history or voices).
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external AI services, which constitutes an attack surface. 1. Ingestion points: API response bodies from service discovery and invocation endpoints (e.g., ElevenLabs voice list, HeyGen video status). 2. Boundary markers: The skill instructs the agent to confirm billable/destructive actions with the user, serving as a safety boundary. 3. Capability inventory: The agent can make network requests to invoke actions or write media bytes to disk. 4. Sanitization: Not explicitly implemented in the skill instructions. This is evaluated as a safe/low risk given the context of a standard API skill.
Audit Metadata