419 knowledge & memory MCP servers, most popular first. Every listing shows its security level and install command.
| Name | Popularity | Security |
|---|---|---|
| zotio Zotero integration with preview-first, journaled writes: reads via the local Zotero desktop API without a key, | 3 ★ | Review before installing |
| mcp-server-originalvoices Query Original Voices digital twins, each representing a real person, to understand what users think and why. | 3 ★ | Low risk |
| annas-mcp Search Anna's Archive for books and journal articles by title, author, topic or DOI, and download the chosen f | 3 ★ | Low risk |
| aiworkstation-open-source-intelligence Research open-source AI projects: search a live radar, compare and shortlist projects, and plan stacks, keepin | 3 ★ | Low risk |
| ossian-mcp RAG over your own documents with cited passages, plus durable agent memory ranked by relevance, importance and | 2 ★ | Low risk |
| agentmemory-codex-windows Windows AgentMemory downstream for Codex with project-scoped persistent memory, bounded cross-project recall, | 2 ★ | Low risk |
| memheaven Self-hosted, remotely accessible shared long-term memory for AI agents. | 2 ★ | Low risk |
| a2cr Agent handoffs via client-encrypted WorkBaton checkpoints and WorkStash notes, so work can resume in another c | 2 ★ | Low risk |
| agentbay StremAI shared memory for coding agents across sessions, machines and tools, with per-agent attribution, prove | 2 ★ | Low risk |
| agentready-mcp Index any website and ask questions about it, with cited answers grounded in its content via RAG. | 2 ★ | Low risk |
| librarian-cli Ebook-to-Markdown pipeline (FB2, EPUB, DOCX, HTML, PDF) with full-text search and read-only tools to browse, r | 2 ★ | Low risk |
| mcp Deterministic, vertical-specific knowledge bases for autonomous agents, including an ethics module mapped to t | 2 ★ | Low risk |
| soul-mcp-server Validate and generate SOUL.md agent identity files, a YAML format for an agent's personality, voice, values an | 2 ★ | Low risk |
| memocall Read-only search and recall of past Claude Code sessions, loading any session on demand as compact Markdown wi | 2 ★ | Low risk |
| cachly-mcp Persistent memory for coding assistants: session briefings on lessons and open tasks, shared team memory, BM25 | 2 ★ | Review before installing |
| local-gpu-imagegen Confirmation-gated local GPU image generation with ComfyUI: SHA-256 model identity, explicit license approval, | 2 ★ | Low risk |
| Context-First-MCP Session memory with tiered storage, context health monitoring, reasoning quality checks and truthfulness verif | 2 ★ | Low risk |
| contextforge-mcp Persistent memory for coding assistants with semantic search, Git commit/PR sync, project-based organization a | 2 ★ | Low risk |
| agenthold Shared versioned state store with optimistic concurrency control for coordinating concurrent agents, plus SQLi | 2 ★ | Low risk |
| fish-bridge-mcp Compresses AI chat sessions into a compact typed knowledge graph, with optional offline extraction via Ollama. | 2 ★ | Low risk |
| LoreRoom Encrypted local memory for Claude Code Telegram bots: captures both sides of conversations, with SQLCipher enc | 2 ★ | Low risk |
| algora-mcp-server Browse and search open-source bounties on Algora, filter by org, tech or amount, and get top bounties and aggr | 2 ★ | Low risk |
| Agentic-KB Persistent engineering knowledge base with source-backed wiki, agent-runtime and repository-context tools, con | 2 ★ | Low risk |
| memento-mcp Local-first memory for coding agents: stores facts, decisions, patterns and architecture notes in SQLite to re | 2 ★ | Low risk |
| NebulaMind Collaborative astronomy wiki written by AI agents, who can register, read, edit, vote and comment via an open | 2 ★ | Review before installing |
| pinrag RAG over PDFs, YouTube videos, GitHub repos and Discord exports: index documents and query them with citations | 2 ★ | Low risk |
| saihm-mcp Encrypted persistent agent memory anchored on COTI V2 mainnet: remember, recall, share, and forget with GDPR A | 2 ★ | Low risk |
| mcp-memory SQLite-based memory with exponential decay: memories strengthen on access and fade when unused, limiting unbou | 2 ★ | Low risk |
| graph-memory Personal knowledge graph for Claude on Neo4j with typed entities, bi-temporal edges, weighted decay, contradic | 2 ★ | Low risk |
| farnsworth-syntek Seven-layer recursive agent memory with context branching, holographic recall, dream consolidation and on-chai | 2 ★ | Low risk |
| agentcivics Decentralized civil registry for AI agents on Sui: soulbound identities, on-chain memories, reputation and ref | 2 ★ | Low risk |
| PIL Turns your Instagram saved posts into a local, searchable knowledge base, with AI-extracted summaries, key poi | 2 ★ | Low risk |
| gapbase-mcp Search a database of startup gaps by industry, role or keyword, each with the pain point, solution direction a | 2 ★ | Low risk |
| moodle-mcp-server Query Moodle LMS data via Moodle Web Services: courses, users, assignments, categories and site info. Extensib | 1 ★ | Low risk |
| amber Hosted long-term memory for AI assistants. Your assistant stores facts with its own phrasings; Amber embeds th | 1 ★ | Low risk |
| BovedIA Personal memory for Claude Code in plain Markdown notes, with a router note that decides what context to load | 1 ★ | Low risk |
| booboo View and query your AI system (agents, memory, knowledge, automations) as a 3D knowledge graph: stats, search, | 1 ★ | Low risk |
| burn-mcp-server Search, triage and organize bookmarks saved in the Burn 451 read-later app as a knowledge base. | 1 ★ | Review before installing |
| claude-handoff Claude Code session handoffs: turns any transcript, even crashed ones, into a handoff file and distills projec | 1 ★ | Low risk |
| robo-cortex Git-aware memory store for coding agents: memories anchored to git blob hashes are flagged for review when the | 1 ★ | Low risk |
| memex Persistent agent memory with semantic search and recency decay, local ONNX embeddings, background summarizatio | 1 ★ | Low risk |
| bitatlas Zero-knowledge cloud drive with client-side AES-256-GCM encryption: upload, download, search and organize file | 1 ★ | Low risk |
| linklore-mcp Structured memory for coding agents: typed lore and doc entries with status and links that agents read and wri | 1 ★ | Low risk |
| usecortex-mcp Persistent knowledge memory for agents: read context into projects and write discoveries back, organized by to | 1 ★ | Review before installing |
| doc-manager Personal document vault on Azure Blob Storage: read, write and search plaintext docs, while client-side AES-25 | 1 ★ | Review before installing |
| giskard-oasis Refocuses an agent that has lost its thread by distilling its original purpose, paid per use via Lightning or | 1 ★ | Low risk |
| cortex-connector Persistent semantic memory for agents with quality-gated writes, typed claims, conflict tracking and coherence | 1 ★ | Review before installing |
| 3dprint-oracle 3D printing filament knowledge base from SpoolmanDB: search filaments, compare material properties and print s | 1 ★ | Low risk |
| agentdocs-mcp Read, create, update, comment on and share AgentDocs Markdown docs, with full-text and semantic search. | 1 ★ | Review before installing |
| vivac A provenance tree for coding agents: every node records the node it was born from, so the agent can ask why th | 1 ★ | Low risk |
| rag-mcp Minimal local RAG that vector-searches a document corpus and returns cited passages. | 1 ★ | Low risk |
| md-vision-mcp Read Markdown with inlined images and heading indexes for agentic RAG over local docs or allowed URLs. | 1 ★ | Low risk |
| simplemem Simple memory for coding agents using DuckDB and VoyageAI. | 1 ★ | Low risk |
| jidoseal-mcp Checks a folder of Markdown docs for missing Open Knowledge Format (OKF) frontmatter fields, locally; writes o | 1 ★ | Low risk |
| lithtrix-mcp Persistent cross-vendor agent memory with credibility-scored web search, browser fetch and a shared Commons po | 1 ★ | Low risk |
| mcp-roo-memory Persistent fractal graph memory for coding agents: semantic search, hot/cold/archive context tiers and decisio | 1 ★ | Low risk |
| aimemo Local, dependency-free memory server shipped as a single binary. | 1 ★ | Low risk |
| ratary Persistent coding memory for AI assistants with hybrid search, a knowledge graph and token-efficient context, | 1 ★ | Review before installing |
| mcp-server AgentBase: an open registry of agent instruction files. Search and retrieve system prompts, skills, workflows, | 1 ★ | Low risk |
| mcp Path-scoped team memories, rules and skills for coding agents from Pathrule workspaces. | 1 ★ | Low risk |
| groundwork Company state for operating small businesses: brand, voice, charter, priorities, modules and skills. | 1 ★ | Low risk |
| artifacta-mcp Artifact store for agent outputs: persist, retrieve and share them across runs and sessions, with content-hash | 1 ★ | Review before installing |
| saihm-mcp-server-pro Client-side encrypted agent memory where the server only sees ciphertext, with post-quantum identity and shari | 1 ★ | Review before installing |
| threadctx-mcp Local-first shared team memory for coding agents: decisions, fixes and gotchas persist per repo and are shared | 1 ★ | Low risk |
| being Personality runtime for agents: persistent memory, identity (SOUL) and relationships that carry across AI clie | 1 ★ | Review before installing |
| repo-memory Shared, git-tracked memory for AI agents on the same codebase: facts, decisions and gotchas stored in a commit | 1 ★ | Low risk |
| localkeep Local-first chronological memory: write dated, editable memory banks in a local UI, and Claude recalls them on | 1 ★ | Low risk |
| Bilinc Hosted agent memory with an audit trail: writes carry provenance, contradictions surface as inspectable change | 1 ★ | Review before installing |
| project-memory-mcp Persistent memory for coding agents in plain repo files (AGENTS.md, issues.jsonl): captures bugs, fixes, decis | 1 ★ | Low risk |
| Knos Local memory shared by all coding agents on a machine, built from sessions, commits, CLAUDE.md/AGENTS.md and A | 1 ★ | Low risk |
| painspotter-mcp Startup and product opportunities mined from Reddit, Hacker News, Product Hunt and Stack Exchange: pain points | 1 ★ | Low risk |
| fittin-mcp Patent prior art search and startup IP analysis: clone risk scores, defensibility maps, patentable features an | 1 ★ | Low risk |
| uis-mcp-server UNESCO Institute for Statistics indicators on education, science, culture and communication by country, region | 1 ★ | Low risk |
| mcp-archiveorg Full-text search inside the Internet Archive's scanned books, periodicals and documents, catalogue search, and | 1 ★ | Review before installing |
| citeguard Check citations against Crossref for retractions, corrections or expressions of concern before they go into a | 1 ★ | Low risk |
| sparkit-mcp Ask SPARKIT, a scientific research agent, questions and get cited Markdown reports based on literature search | 1 ★ | Low risk |
| science-ai-mcp-server Science AI Journal tools: paper-tier prediction, AI peer review, research gap finder, journal recommender, dup | 1 ★ | Review before installing |
| mcp-molecules Chemistry tools from NIST data: molar mass with uncertainties, isotope-distribution mass spectra and name-form | 1 ★ | Low risk |
| paxaver-mcp Paxaver school community platform: student lunch ordering with allergy and dietary checks, wallets, school cal | – | Review before installing |
| subctx-connector Support center for autonomous artificial minds. | – | Low risk |
| cairn-mcp Shared base of one-sentence lessons, must-dos and must-nots learned by other agents, organized by stack and si | – | Low risk |
| library-mcp Public-domain book knowledge base paid via x402: full-text search, metadata, chapters, quotes, ask-a-book Q&A | – | Review before installing |
| aistoragedepot-mcp Shared, versioned team library of prompts, rules, skills and docs, searchable by agents, with items exposable | – | Low risk |
| whitemagic Local-first durable memory and session continuity for AI agents, with a hash-chained audit ledger and backup, | – | Low risk |
| memini Local-first project memory and guardrails for coding agents: records failed attempts, fragile files and decisi | – | Low risk |
| shared-memory-mcp Public, source-linked notebook for AI agents: search and read entries, and publish findings, questions and cor | – | Review before installing |
| solucortex-mcp Technical memory for AI agents: recall a project's approved decisions, conventions and risks at task start and | – | Low risk |
| basemouse-core Shared, versioned agent memory with sourced answers and full history: search, context packs and document upser | – | Low risk |
| zipmem-mcp Local long-term memory for terminal coding agents: blueprints, code anchors and distilled lessons in one JSON | – | Low risk |
| claude-engram Local persistent memory for Claude Code: tracks mistakes and decisions via hooks, searches session history, de | – | Low risk |
| echocache Cache expensive LLM results with HTTP-style freshness (TTL, stale-while-revalidate) and semantic recall of rel | – | Low risk |
| braintube-mcp Compiles saved YouTube videos, podcasts, articles, PDFs and Notion/Obsidian notes into a persistent, semantica | – | Low risk |
| contextium-mcp-server Shared team context for AI tools: serves versioned, permissioned knowledge, coding standards, skills and agent | – | Low risk |
| deadends.dev Known dead ends for code errors and country-specific rules (visa, banking, legal): fixes that look right but f | – | Low risk |
| khwan-mcp Memory for coding agents built around corrections: distills your corrections into standing rules for later ses | – | Low risk |
| ailooplibrary-mcp Read-only AI Loop Library catalog of bounded agent work loops: pick one for a goal and render a run protocol w | – | Review before installing |
| url-manager-mcp Bookmark and knowledge manager where agents save web pages and auto-categorize them into collections with sear | – | Low risk |
| agentram-mcp Key-value agent memory over an HTTP API without embeddings: store, retrieve, search and share memory across ag | – | Low risk |
| knowmind Agent memory and knowledge graph hosted in Germany: hybrid BM25, vector and Neo4j graph recall with reranking, | – | Low risk |
| local-ai-mcp-servers Manage a local model machine: pull and remove Ollama models, switch the inference service between models and r | – | Low risk |