convex-agent
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for agents to process untrusted data from message history and RAG search results. (1) Ingestion points: The agent ingests data from persisted message threads and vector search results as specified in SKILL.md. (2) Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions to isolate retrieved documents from the system prompt. (3) Capability inventory: The agent supports tool calls, vector search retrieval, and direct model invocation. (4) Sanitization: No procedures are described for sanitizing or filtering content retrieved from the vector database before it is processed by the model.
Audit Metadata