architect-structured-procedures
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill exclusively targets the official ElevenLabs API domain (api.elevenlabs.io) for managing agent procedures.
- [SAFE]: Sensitive information is handled securely using the $API_KEY placeholder, preventing hardcoded credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill allows the agent to process procedure content retrieved from the ElevenLabs API, which represents an attack surface for indirect instructions.
- Ingestion points: Existing procedure content is retrieved via the
GET /v1/convai/agents/$AGENT_ID/branches/$BRANCH_ID/procedures/$PROCEDURE_IDendpoint. - Boundary markers: No explicit delimiters or instructions to ignore embedded content are provided in the skill instructions.
- Capability inventory: The skill uses REST API actions (
GET,POST,PATCH) to manage agent state on the ElevenLabs platform. It does not perform file system operations or execute arbitrary shell commands. - Sanitization: No input validation or sanitization of the retrieved procedure content is specified before the agent processes or converts it.
- Note: The risk is considered minimal as these are standard administrative operations within the vendor's own ecosystem.
Audit Metadata