Qdrant
Semantic memory layer built on the Qdrant vector search engine.
🛡️ AgentReady threat assessment
MAESTRO 7-layer threat model + OWASP AIVSS risk score for Qdrant, derived from its capabilities.
AIVSS 8.6 · High
View MAESTRO 7-layer threat model →Overview
Qdrant's official MCP server implements a semantic memory layer, letting agents store and retrieve embeddings for RAG. Retrieved chunks are fed back into the model, so poisoned vectors are an indirect prompt-injection route worth auditing.
Key features
- Vector store and retrieve
- Semantic memory for RAG
- Collection management
Use cases
- Add long-term semantic memory
- Power a RAG pipeline