agent-openai-memory
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the
databricks-openaiPython package. This is a vendor-official library used to integrate session memory capabilities. - [INDIRECT_PROMPT_INJECTION]: The skill implements a session memory system that ingests and persists user conversation history. This creates an attack surface where malicious instructions could be stored and later re-processed by the agent.
- Ingestion points: Conversation messages and session IDs are processed via the
agent_server/agent.pyentry point. - Boundary markers: The provided code snippets do not explicitly show the use of delimiters or instructions to prevent the agent from executing commands found within the retrieved history.
- Capability inventory: The agent communicates with Databricks Lakebase endpoints via network requests to perform session storage and retrieval operations.
- Sanitization: The implementation does not demonstrate explicit validation or filtering of the stored conversation content.
Audit Metadata