{"id":"w3-mcp-server-qdrant","name":"w3-mcp-server-qdrant","homepage":"https://pypi.org/project/w3-mcp-server-qdrant/","repo_url":"https://github.com/famtong8-dev/w3-mcp-server-qdrant","category":"ai-ml","subcategories":[],"tags":["mcp","qdrant","vector-search","ollama","embeddings","python","retrieval","reranking","hyde","rrf"],"what_it_does":"Provides a Python MCP server exposing tools to query a Qdrant vector database using text queries. It can auto-generate embeddings via an external Ollama server and supports advanced retrieval techniques such as query expansion (with RRF merging), HyDE, and LLM-based reranking.","use_cases":["Semantic search over Qdrant collections from an MCP-capable client (e.g., Claude Code)","Improving recall/precision for retrieval with query expansion, HyDE, and reranking","Listing and inspecting Qdrant collections and basic metadata for retrieval setup","Building agent workflows that require vector search with structured (JSON/Markdown) outputs"],"not_for":["Production deployments needing a hardened remote HTTP API surface (the interface is primarily MCP stdio)","Workloads requiring strict data residency/compliance guarantees without additional platform controls","Use cases requiring management operations beyond search/list (create/update/delete collections/points are not described)"],"best_when":"You want an agent-callable vector search tool that enriches queries via Ollama-generated embeddings and optionally uses additional LLM steps (expansion/rerank) to improve retrieval quality.","avoid_when":"You cannot run/allow the supporting services (Qdrant and Ollama) or you need robust built-in auth/rate limiting controls at the MCP server boundary.","alternatives":["Qdrant’s own HTTP API (direct integration)","Open-source MCP vector search servers for other vector databases (if available)","Using a dedicated RAG/retrieval service (hosted) that wraps embeddings + vector search with standard auth and observability"],"af_score":55.8,"security_score":42.2,"reliability_score":21.2,"package_type":"mcp_server","discovery_source":["pypi"],"priority":"low","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-04-04T21:39:53.021841+00:00","interface":{"has_rest_api":false,"has_graphql":false,"has_grpc":false,"has_mcp_server":true,"mcp_server_url":null,"has_sdk":false,"sdk_languages":[],"openapi_spec_url":null,"webhooks":false},"auth":{"methods":["MCP stdio (stdio transport implied by usage examples)","Qdrant API key via QDRANT_API_KEY (optional per README)"],"oauth":false,"scopes":false,"notes":"No MCP-specific authentication or authorization mechanism is described in the README. Qdrant authentication is mentioned as optional via QDRANT_API_KEY."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"Pricing is not described; costs would primarily be external (self-hosted Qdrant/Ollama and any LLM usage for reranking/expansion)."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":[],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":55.8,"security_score":42.2,"reliability_score":21.2,"mcp_server_quality":78.0,"documentation_accuracy":72.0,"error_message_quality":0.0,"error_message_notes":null,"auth_complexity":80.0,"rate_limit_clarity":5.0,"tls_enforcement":25.0,"auth_strength":45.0,"scope_granularity":20.0,"dependency_hygiene":60.0,"secret_handling":65.0,"security_notes":"Security is mostly delegated to the underlying deployments: Qdrant URL/optional API key and Ollama base URL. The README does not describe TLS requirements, MCP authentication, request validation/auditing, or rate limiting. Secret handling best practices are not evidenced in the provided text (no mention of logging/redaction), though using environment variables for keys is implied.","uptime_documented":0.0,"version_stability":35.0,"breaking_changes_history":30.0,"error_recovery":20.0,"idempotency_support":"false","idempotency_notes":null,"pagination_style":"none","retry_guidance_documented":false,"known_agent_gotchas":["Advanced options (expand_query/use_hyde/rerank) may significantly increase latency since they require additional LLM calls and parallel searches.","If Ollama or Qdrant is unavailable, the server may hang or fail; README suggests manual health checks but does not describe MCP-level retry semantics.","No pagination mechanism is described; limit is a single parameter (no continuation token).","Reranking/query-expansion depend on OLLAMA_RERANK_MODEL being available; misconfiguration will break those paths."]}}