iblai-api-agent-session
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a "metadata pass-through" feature where arbitrary JSON content is appended to the agent's prompt.
- Ingestion points: The
metadataobject in the JSON payload for the/api/agent/chat/endpoint and thepostMessageAPI used by embedded iframes to update context. - Boundary markers: Injected metadata is wrapped in
<CONTEXT METADATA>and</END CONTEXT METADATA>tags within the prompt. - Capability inventory: The skill performs network operations against vendor-owned APIs for chat streaming and session management (
*.iblai.app). - Sanitization: There is no evidence of validation or sanitization for the keys or values within the metadata object before it is interpolated into the prompt.
- [COMMAND_EXECUTION]: The documentation provides
curlcommand examples for interacting with session and history APIs, involving the use of environment variables for authentication tokens.
Audit Metadata