Shared, encrypted, verifiable team memory for coding agents (MCP server)
MCP (Model Context Protocol) Server for Retrieval Augmented Generation using Mongo and Gemini, OpenAI, with A2A support Grok
Stable MCP supervisor proxy for stdio servers: cached discovery, unified mcp.wrapper lifecycle control, and on-demand backend processes.
Model Context Protocol (MCP) server for FalkorDB — read-only graph access for AI assistants
Personal knowledge system — MCP server with PostgreSQL semantic search and temporal knowledge graph
An MCP server to build knowledge graph of OWL ontologies
a mcp server for interacting with ChromaDB
A minimal Rust vector-memory server that works with Qdrant and Ollama.
Rust MCP server for HaloPSA - ticket management, workflow actions, and semantic search via the Model Context Protocol
MCP server to allow agents to learn from my personal knowledge management system.
MCP server for searching, indexing, and retrieving scientific papers across arXiv, INSPIRE-HEP, Semantic Scholar, OpenAlex, CrossRef, NASA ADS, Europe PMC, DOAJ, and viXra. Local hybrid search with BM
Apple ecosystem MCP server for Claude Code, Cursor, and other AI assistants — Notes, Reminders, Contacts with BERT semantic search
Local-first semantic-search MCP server over a personal CS/FP/graphics library
Semantic code-search MCP server for AI coding agents: hybrid retrieval (embeddings + BM25 + reranking), tree-sitter AST chunking, 30-100x fewer tokens
CogniMem is a local MCP server in Rust that gives coding agents a persistent cognitive memory layer.
Persistent semantic memory service and MCP adapter for AI agents, powered by PostgreSQL and local Ollama models.
A practical repository focused on mastering Spring AI concepts through hands-on implementations including RAG pipelines, vector search, AI agents, tool calling, memory handling, MCP servers, embedding
A high-performance Model Context Protocol (MCP) server written in Rust that extends AI models into autonomous coding agents with semantic memory, built-in Git tools, and codebase indexers
CodeNinja Atoms: a sovereign AI operating system for the physical world. Open reference architectures for physical AI, with ontologies, model registers, costs and DOIs. CC BY 4.0.
Layered, self-learning knowledge base for teams, served over the Model Context Protocol. PostgreSQL + pgvector hybrid search, roles, audit log, YAML layer presets.
Parallax - an MCP server of independent correctives for a model's predictable failure modes: adversarial verification, deterministic checks (Z3/evalexpr), verified cross-session memory, cited research
Serveur MCP pour bases de connaissance markdown versionnées en git. Des sessions Claude isolées les lisent et y déposent des propositions ; seul le rôle gestionnaire intègre.
Long-term memory for coding agents over MCP: hybrid search over Markdown notes with a link graph, on PostgreSQL, Qdrant, and Ollama
(BACKEND )An MCP-compatible vector search server built in Go, powered by Pinecone and OpenAI.
Best Token-Efficient AI Coding Tools 2026 – Reduce Costs 90% with Local Memory & MCP
Cross-session persistent memory MCP server for AI assistants — local SQLite + FTS5, zero external dependencies
MCP server for remote management of project memory banks
Long-term memory for AI agents. Pass one sentence, and it remembers, recalls, or forgets.
Geospatial AI MCP server — satellite imagery, embeddings, weather, GNS governance
An MCP memory server. One memory your agents share — across models, devices and apps.
Persistent memory and knowledge graphs for AI agents. Hybrid search, context checkpoints, and more.
Typed on-prem knowledge graph for AI agents — read-only for humans, write-only for agents via MCP.
OpenContext memory storage + search SDK with HTTP and MCP server entry points.
MCP server for semantic search using local Qdrant and Ollama (default) with support for OpenAI, Cohere, and Voyage AI
Local visual UI cache for AI agents using perceptual hashing, CLIP, and 3D spatial grounding.
Private on-device memory & retrieval for AI assistants — offline vector + lexical search.
ディレクトリ全体
RAG、グラフ、エージェントの長期メモリ。 2,618 件のサーバーを利用の多い順に掲載しています。それぞれ公開元のページが開きます。