20 Best Vector Database MCP Servers, Compared (November 2026)

The short answer

Which one, for your situation

Top MCP servers for Vector Database

The ranking, with the evidence

• 1

  mcp-server-qdrant

  Steady

  qdrant/mcp-server-qdrant ↗

  Best for: Persistent semantic memory for an LLM workflow: it exposes qdrant-store and qdrant-find to save and retrieve memories in Qdrant.

  It exposes two tools, qdrant-store for saving information with optional metadata and qdrant-find for retrieving relevant stored information from a Qdrant collection. Before choosing it, note that configuration requires either QDRANT_URL or QDRANT_LOCAL_PATH (not both) and a collection name unless a default collection is set.

  Score 69.9: show every number behind it

  • Adoption80 / 100 · weight 40%
    • GitHub stars80
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance57 / 100 · weight 24%
    • Last commit touching this server88
    • Repository maintenance grade30
    • Commit cadence5
  • Momentum46 / 100 · weight 14%
    • Stars gained, relative to size43
    • Stars gained, absolute52
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality68 / 100 · weight 13%
    • Tool description quality60
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  69.9

  × relevance: the keyword is dedicated here

  1.00

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  69.9

• 2

  MCP Memory Service

  Community favourite

  doobidoo/mcp-memory-service ↗

  Best for: Adding persistent semantic memory to Claude Desktop: it exposes store_memory, retrieve_memory, and search_by_tag backed by a self-hosted ChromaDB service.

  MCP Memory Service exposes three tools, store_memory, retrieve_memory, and search_by_tag, and documents MCP and REST API interfaces for storing and retrieving memory. Choosing it means self-hosting the service and relying on ChromaDB and sentence transformers for local embeddings rather than a hosted memory backend.

  Score 67.2: show every number behind it

  • Adoption88 / 100 · weight 40%
    • GitHub stars83
    • PyPI downloads33
      PyPI downloads show no weekday rhythm; halved
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum82 / 100 · weight 14%
    • Stars gained, relative to size85
    • Stars gained, absolute80
    • PyPI download trend79
  • Tool quality50 / 100 · weight 13%
    • Tool description quality50
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  86.2

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  67.2

• 3

  Haiku RAG

  Community favourite

  ggozad/haiku.rag ↗

  Best for: Local document Q&A with citations: it ingests PDFs, office files, HTML, Markdown and images, and provides MCP search, reading and a Python sandbox.

  Haiku RAG ingests PDFs, office documents, HTML, Markdown and images into LanceDB and documents a CLI, Python API, and MCP server for search, document reading and a Python sandbox over the documents. It requires Python 3.12 or newer and defaults to Ollama for embeddings and answers, though other Pydantic AI providers can be configured.

  Score 65.7: show every number behind it

  • Adoption80 / 100 · weight 40%
    • GitHub stars70
    • PyPI downloads65
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum75 / 100 · weight 14%
    • Stars gained, relative to size76
    • Stars gained, absolute58
    • PyPI download trend91
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  84.2

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  65.7

• 4

  Pinecone Developer MCP

  Community favourite

  pinecone-io/pinecone-mcp ↗

  Best for: Developers configuring and querying Pinecone indexes from Cursor or Claude: it exposes 12 tools for documentation search, index setup, upserts, and search.

  The Pinecone Developer MCP Server connects AI assistants to Pinecone through 12 tools for searching official documentation, listing and describing indexes, creating indexes, upserting documents and records, searching indexes, reranking, and cascading search. Index management and querying require a Pinecone API key; without one, the server can still search documentation but cannot manage or query indexes.

  Score 64.8: show every number behind it

  • Adoption79 / 100 · weight 40%
    • GitHub stars47
    • npm downloads72
      downloads capped: not corroborated by repository audience
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance77 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade55
    • Commit cadence40
  • Momentum42 / 100 · weight 14%
    • Stars gained, relative to size29
    • Stars gained, absolute13
    • npm download trend100
  • Tool quality88 / 100 · weight 13%
    • Tool description quality80
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  76.2

  × relevance: the keyword is dedicated here

  1.00

  × continuity: actively changing

  1.00

  × evidence: modest but real audience

  0.85

  Composite score

  64.8

• 5

  XERJ

  Community favourite

  xerj-org/xerj ↗

  Best for: Coding agents that need local repository search: it indexes code and docs and returns cited file:line passages across BM25, vector, hybrid, and batch queries.

  XERJ is a local Rust binary and MCP server that indexes code, docs, logs, JSONL/CSV and PDFs and returns cited passages through CLI commands, an Elasticsearch-compatible API, and MCP integration, including BM25, vector, hybrid and batch search plus agent memory namespaces. Before choosing it, note that it runs as a local binary with an explicit data directory, and neural embeddings are opt-in and require an additional download.

  Score 63.8: show every number behind it

  • Adoption100 / 100 · weight 40%
    • GitHub stars100
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum≈0 / 100 · weight 14%
    • Stars gained, relative to sizenot measured
      no snapshot history yet
    • Stars gained, absolutenot measured
      no snapshot history yet
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  81.8

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  63.8

• 6

  MCP Server for Milvus

  Steady

  zilliztech/mcp-server-milvus ↗

  Best for: Connecting Claude Desktop or Cursor to an existing Milvus instance over stdio, SSE, or Streamable HTTP: the README documents setup for all three transports.

  MCP Server for Milvus provides an MCP interface to a running Milvus instance and documents stdio, SSE, and Streamable HTTP transports, with configuration examples for Claude Desktop and Cursor. No PyPI package is attributed and no npm package is published, so setup follows the repository's uv commands and requires a running Milvus instance plus Python 3.10 or higher.

  Score 61.3: show every number behind it

  • Adoption60 / 100 · weight 40%
    • GitHub stars60
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance64 / 100 · weight 24%
    • Last commit touching this server88
    • Repository maintenance grade30
    • Commit cadence40
  • Momentum30 / 100 · weight 14%
    • Stars gained, relative to size35
    • Stars gained, absolute24
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  61.3

  × relevance: the keyword is dedicated here

  1.00

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  61.3

• 7

  Remnic

  Emerging

  joshuaswarren/remnic ↗

  Best for: When multiple MCP clients need one local memory store: Remnic documents shared markdown/YAML records on disk plus local JSON and SQLite state.

  Remnic documents a local-first memory layer for MCP and other agent clients, storing memory as markdown files with YAML frontmatter plus local JSON and SQLite state and offering hybrid BM25 and vector recall with per-result provenance. The measured package status is that no npm package is published, and the exposed MCP tools were not measured, so installation and tool behavior require checking the repository.

  Score 58.7: show every number behind it

  • Adoption59 / 100 · weight 40%
    • GitHub stars59
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum70 / 100 · weight 14%
    • Stars gained, relative to size84
    • Stars gained, absolute49
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  75.3

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  58.7

• 8

  Synaplan

  Community favourite

  metadist/synaplan ↗

  Best for: Self-hosted teams needing document search and persistent memories via MCP: it exposes Qdrant and Apache Tika RAG plus long-term memories as MCP tools and resources.

  Synaplan is a self-hosted AI workspace that provides semantic document search through Qdrant and Apache Tika, long-term user memories, and multi-provider AI chat as MCP tools and resources. It requires running a PHP/Symfony stack, and its MCP tool list has never been inspected, so confirm the available calls before depending on them.

  Score 54.2: show every number behind it

  • Adoption69 / 100 · weight 40%
    • GitHub stars69
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum≈0 / 100 · weight 14%
    • Stars gained, relative to sizenot measured
      no snapshot history yet
    • Stars gained, absolutenot measured
      no snapshot history yet
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  69.5

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  54.2

• 9

  cortexdb

  Community favourite

  liliang-cn/cortexdb ↗

  Best for: Several AI agents needing shared memory without an embedding model: it documents a gRPC shared-brain server and plugins backed by one SQLite file.

  CortexDB documents a Go library, Claude Code and Codex plugins, an Obsidian plugin, and a shared-brain gRPC server, with a SQLite-backed brain for vector and keyword search, scoped agent memory, RAG context packs, and RDF/SPARQL plus read-only Cypher graph queries. Before choosing it, note that no npm package is published and no PyPI package is attributed, so installation should follow the documented Go library, plugin, or gRPC server routes.

  Score 53.5: show every number behind it

  • Adoption67 / 100 · weight 40%
    • GitHub stars67
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum≈0 / 100 · weight 14%
    • Stars gained, relative to sizenot measured
      no snapshot history yet
    • Stars gained, absolutenot measured
      no snapshot history yet
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  68.6

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  53.5

• 10

  SocratiCode

  Community favourite

  giancarloerra/SocratiCode ↗

  Best for: Working in a large, unfamiliar codebase where code must stay local: it indexes and searches code, builds dependency graphs, and supports impact and flow analysis.

  SocratiCode ships 26 MCP tools for indexing and watching codebases, semantic search, dependency graph construction and queries, impact analysis, symbol inspection, and context-artifact search. It assumes Docker, a Qdrant container, Ollama, and an embedding model, and it needs an MCP host that supports local stdio servers.

  Score 53.3: show every number behind it

  • Adoption99 / 100 · weight 40%
    • GitHub stars88
    • npm downloads70
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum47 / 100 · weight 14%
    • Stars gained, relative to size44
    • Stars gained, absolute66
    • npm download trend27
  • Tool quality76 / 100 · weight 13%
    • Tool description quality68
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  88.8

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  53.3

• 11

  CodeGraph CLI MCP Server

  Community favourite

  Jakedismo/codegraph-rust ↗

  Best for: Indexing a multi-language codebase for AI assistants that need semantic search plus relationship context: it builds a knowledge graph with embeddings and tiered indexing.

  CodeGraph documents a CLI that indexes a codebase into a graph with vector embeddings and supports commands such as codegraph init, codegraph index, codegraph estimate, and codegraph agent context, with fast, balanced, and full indexing tiers. The balanced and full tiers require external language servers for the target languages, and indexing fails fast when those required tools are missing.

  Score 53.2: show every number behind it

  • Adoption74 / 100 · weight 40%
    • GitHub stars74
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance89 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade55
    • Commit cadence100
  • Momentum60 / 100 · weight 14%
    • Stars gained, relative to size63
    • Stars gained, absolute57
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust0 / 100 · weight 9%
    • License0
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  68.2

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  53.2

• 12

  TxtAI MCP Server

  Community favourite

  neuml/txtai ↗

  Best for: Teams already using txtai for vector search and wanting MCP access: the README documents an MCP API alongside the web API.

  It exposes txtai's embeddings database and LLM workflow capabilities through the MCP API documented in the README, with the same functionality also available via the web API. Before choosing it, note that no npm package is published and no PyPI package is attributed, so there is no measured package-registry distribution to install from.

  Score 53.1: show every number behind it

  • Adoption100 / 100 · weight 40%
    • GitHub stars100
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum48 / 100 · weight 14%
    • Stars gained, relative to size30
    • Stars gained, absolute75
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  88.5

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  53.1

• 13

  claude-skills-mcp

  Abandoned but popular

  K-Dense-AI/claude-skills-mcp ↗

  Best for: Starting a new task and needing relevant Claude Agent Skills: it provides semantic skill search, document retrieval, and a complete inventory listing.

  It exposes three tools for semantic search over Claude Agent Skills, reading specific skill documents, and listing the loaded skill inventory. It runs from a Python package via uvx, and the first run auto-downloads and starts a separate backend in the background.

  Score 52.5: show every number behind it

  • Adoption65 / 100 · weight 40%
    • GitHub stars65
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance63 / 100 · weight 24%
    • Last commit touching this server88
    • Repository maintenance grade55
    • Commit cadence5
  • Momentum42 / 100 · weight 14%
    • Stars gained, relative to size46
    • Stars gained, absolute37
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality85 / 100 · weight 13%
    • Tool description quality77
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  67.3

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  52.5

• 14

  Local RAG

  Community favourite

  shinpr/mcp-local-rag ↗

  Best for: Searching private documents without sending them to an embedding API: it indexes PDF, DOCX, TXT, and Markdown files locally and searches them offline.

  It exposes nine tools covering sync and sync status, single-file and in-memory ingestion, hybrid keyword plus semantic search through query_documents, reading chunk neighbors for surrounding context, file listing, deletion, and index status. Before choosing it, note that it requires Node.js 22.20 or later and internet access on first use to download the package and the embedding model, and that Excel, PowerPoint, standalone images, and source-code files fall outside its ingestion scope.

  Score 52.4: show every number behind it

  • Adoption87 / 100 · weight 40%
    • GitHub stars66
    • npm downloads77
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum58 / 100 · weight 14%
    • Stars gained, relative to size77
    • Stars gained, absolute53
    • npm download trend29
  • Tool quality88 / 100 · weight 13%
    • Tool description quality80
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  87.3

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  52.4

• 15

  ArcadeDB Multi-Model DBMS

  Community favourite

  ArcadeData/arcadedb ↗

  Best for: Managing vector-enabled ArcadeDB deployments through MCP: it exposes 10 tools for database listing, schema, queries, writes, server status, profiling, and settings.

  ArcadeDB Multi-Model DBMS exposes 10 MCP tools for listing databases, reading schema, running read-only queries, executing write commands, checking server status, profiling queries, and getting or setting server configuration. Before choosing it, note that it operates against an ArcadeDB server and scopes database listing to the authenticated user.

  Score 50.9: show every number behind it

  • Adoption77 / 100 · weight 40%
    • GitHub stars77
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum75 / 100 · weight 14%
    • Stars gained, relative to size79
    • Stars gained, absolute69
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality81 / 100 · weight 13%
    • Tool description quality73
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  84.8

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  50.9

• 16

  mcp-vector-search

  Abandoned but popular

  bobmatnyc/mcp-vector-search ↗

  Best for: Local semantic search over a codebase from an AI assistant: the README documents a command-line tool and MCP integration that index the project.

  The README documents a command-line tool and MCP integration that index a local codebase for semantic search, with LanceDB as the default backend. Before choosing it, note that installation requires Python 3.11 or newer and the mcp-vector-search PyPI package; the README recommends running mcp-vector-search setup afterward.

  Score 50.2: show every number behind it

  • Adoption63 / 100 · weight 40%
    • GitHub stars44
    • PyPI downloads56
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance46 / 100 · weight 24%
    • Last commit touching this server68
    • Repository maintenance grade30
    • Commit cadence5
  • Momentum38 / 100 · weight 14%
    • Stars gained, relative to size48
    • Stars gained, absolute24
    • PyPI download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust0 / 100 · weight 9%
    • License0
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  50.2

  × relevance: the keyword is dedicated here

  1.00

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  50.2

• 17

  second-brain-cloudflare

  Community favourite

  rahilp/second-brain-cloudflare ↗

  Best for: Using multiple MCP AI clients across various sessions: it provides a persistent, self-hosted memory system that recalls context based on meaning.

  The server functions as a semantic memory layer, allowing users to capture notes and recall context using natural language through REST or Model Context Protocol interfaces. It operates entirely within a Cloudflare account and depends on Cloudflare Workers, D1, and Vectorize for its operation.

  Score 50.0: show every number behind it

  • Adoption73 / 100 · weight 40%
    • GitHub stars73
    • npm downloadsnot measured
      npm names no repository for second-brain, so its downloads cannot be attributed
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum88 / 100 · weight 14%
    • Stars gained, relative to size100
    • Stars gained, absolute71
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  83.3

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  50.0

• 18

  MCP MariaDB Server

  Abandoned but popular

  MariaDB/mcp ↗

  Best for: Semantic search inside an existing MariaDB instance: it creates vector stores, inserts documents, and runs embedding-based similarity search alongside read-only SQL.

  The server exposes database tools for listing databases and tables, retrieving table schemas with or without foreign key relations, creating databases, and running read-only SELECT, SHOW, and DESCRIBE queries, alongside vector store tools for creating and deleting stores, inserting documents, and running similarity search. Its vector tools stay disabled unless an embedding provider (OpenAI, Gemini, or HuggingFace) is configured with the matching API key, and since no npm or PyPI package is published, it has to be installed from the repository.

  Score 49.9: show every number behind it

  • Adoption58 / 100 · weight 40%
    • GitHub stars58
    • npm downloadsnot measured
      no npm package
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance57 / 100 · weight 24%
    • Last commit touching this server100
      dated from the last commit on the default branch, re-read from GitHub at publication; the stored date would have published 87 days ago
    • Repository maintenance grade5
    • Commit cadence5
  • Momentum65 / 100 · weight 14%
    • Stars gained, relative to size77
    • Stars gained, absolute46
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integrates100

  Weighted mean of the five

  64.0

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  49.9

• 19

  knowledge-rag

  Community favourite

  lyonzin/knowledge-rag ↗

  Best for: Local RAG over a document set for Claude Code or other MCP clients, using hybrid semantic and BM25 search with cross-encoder reranking.

  The server exposes 13 MCP tools for searching, adding, updating, removing, listing, reindexing, and evaluating retrieval quality over a local document knowledge base, with hybrid semantic and BM25 search plus cross-encoder reranking. Setup requires a first model download for in-process FastEmbed ONNX embeddings, and add_from_url makes outbound HTTP requests.

  Score 49.6: show every number behind it

  • Adoption70 / 100 · weight 40%
    • GitHub stars62
    • PyPI downloads53
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance100 / 100 · weight 24%
    • Last commit touching this server100
    • Repository maintenance grade100
    • Commit cadence100
  • Momentum63 / 100 · weight 14%
    • Stars gained, relative to size61
    • Stars gained, absolute42
    • PyPI download trend89
  • Tool quality100 / 100 · weight 13%
    • Tool description quality98
    • Built and inspected by Glama100
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  82.7

  × relevance: the keyword is tagged here

  0.60

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  49.6

• 20

  Design Systems MCP Server

  Steady

  southleft/design-systems-mcp ↗

  Best for: When an AI client needs dated, cited design-system and agent-interface protocol answers: it provides semantic vector search over curated entries with source reliability badges.

  It exposes a public MCP endpoint at https://design-systems-mcp.southleft.com/mcp and provides vector plus keyword search over curated design-system knowledge, using Supabase pgvector and a Postgres full-text path for when embeddings are unavailable. Before choosing it, note that local development assumes Supabase pgvector, an embeddings provider, and credentials in .dev.vars, while the hosted endpoint requires no installation.

  Score 49.5: show every number behind it

  • Adoption59 / 100 · weight 40%
    • GitHub stars59
    • npm downloadsnot measured
      npm names no repository for design-systems-mcp, so its downloads cannot be attributed
    • Used through Glamanot measured
      not used through Glama in the last 30 days
  • Maintenance64 / 100 · weight 24%
    • Last commit touching this server88
    • Repository maintenance grade30
    • Commit cadence40
  • Momentum49 / 100 · weight 14%
    • Stars gained, relative to size57
    • Stars gained, absolute37
    • npm download trendnot measured
      no download history for the selected registry
  • Tool quality≈68 / 100 · weight 13%
    • Tool description quality≈68
      tool descriptions not yet scored
    • Built and inspected by Glamanot measured
      never built and inspected by Glama
  • Trust100 / 100 · weight 9%
    • License100
    • Published by the vendor it integratesnot measured
      not published by the vendor it integrates

  Weighted mean of the five

  63.5

  × relevance: the keyword is declared here

  0.78

  × continuity: actively changing

  1.00

  × evidence: widely adopted

  1.00

  Composite score

  49.5

Questions people ask

Other comparisons

• Best Figma MCP servers
• Best Obsidian MCP servers
• Best YouTube MCP servers
• Best Jira MCP servers
• Best PostgreSQL MCP servers
• Best MySQL MCP servers
• Best GitHub MCP servers
• Best Notion MCP servers
• Best Linear MCP servers
• Best Discord MCP servers
• Best MongoDB MCP servers
• Best CRM MCP servers
• Best Payments MCP servers
• Best Ecommerce MCP servers
• Best Social Media MCP servers
• Best Browser Automation MCP servers
• Best Email MCP servers
• Best PDF MCP servers
• Best Kubernetes MCP servers
• Best Docker MCP servers
• Best AWS MCP servers
• Best Spreadsheets MCP servers
• Best SEO MCP servers
• Best Web Search MCP servers
• Best Web Scraping MCP servers
• Best Code Execution MCP servers
• Best Calendar MCP servers
• Best Observability MCP servers
• Best Image Generation MCP servers
• Best Speech & Transcription MCP servers
• Best Maps MCP servers
• Best Travel MCP servers
• Best Stock Market MCP servers
• Best Web Analytics MCP servers
• Best Memory MCP servers
• Best Documentation MCP servers
• Best Security MCP servers