Notes & Knowledge MCP servers

Local-first personal memory that distills Feishu/Lark chats and Claude Code/Codex sessions into a private knowledge graph of people, projects and tasks.

MCP server for AI agent profiles and smart notes. 60+ coding prompt packs with expert personas.

Work with OWL/TTL/RDF ontologies: SPARQL queries, HermiT/Pellet reasoning and CRUD operations on knowledge graphs.

Read and write a self-hosted TriliumNext knowledge base over its ETAPI.

Inspectable memory for Claude Code: read-only search and session compaction over a local, signal-scored parse of its project logs, with no outbound calls.

Persistent schema memory for PostgreSQL, MySQL, MSSQL and SQLite: table and column annotations, a foreign-key graph and searchable notes. Read-only.

Explore Claude Code's source as queryable architecture modules: trace cross-cutting concerns, search patterns and read source code locally.

Offline semantic search over library documentation using a pre-built, version-pinned vector index fetched and verified on first launch.

InvGate Service Desk: incidents, users, knowledge base, assets, custom fields, workflows and time tracking, read-only by default with opt-in writes.

Self-hosted Markdown memory for AI agents: notes you can read and edit, with hybrid search.

LEGO sets, parts, minifigs and inventories from Rebrickable: search and compare sets, find parts by colour, browse themes and discover alternate builds.

Local-first governed shared memory for coding agents with CLI and desktop console: classify, dedupe, supersede, quarantine and roll back memories.

Academic paper knowledge base over local PDFs: MinerU parsing, metadata normalization, Qdrant vector search with reranking and paginated full-text reading.

Portable, privacy-first AI identity: a home directory for your agents with profile management, skills, resume import and team sync.

Search, read and act on Apple Notes — folders, attachments, writes off by default

ComfyUI MCP server: knowledge base, workflow builder, real-time execution, models, plugins.

Interact with Confluence Data Center via its REST API.

Google Docs MCP server with full tab support, markdown extraction, and batch updates.

Bilingual (English/Spanish) dual-memory system on PostgreSQL, pgvector and NVIDIA NIM with progressive retrieval, deduplication, checkpoints and auto-repair.

Create, search and manage Knowtis collaborative notes from AI assistants.

Persistent knowledge layer for agents: knowledge base, investigation threads and a journal with multi-agent attribution, on SQLite, PostgreSQL or Supabase.

Private local research second brain: project-aware memory, cited answers from your library, linked notes and meetings, daily briefs and a meeting assistant.

Unofficial Notion access via the private API (token_v2): read and write across your whole workspace without an integration token or per-page sharing.

Xiaohongshu (rednote / 小红书) MCP server for Claude Desktop, Cursor and AI agents — search, read, comment, like and publish notes via npx. Node.js + Playwright port of xiaohongshu-mcp.

YouTube knowledge base: summarize videos, fetch transcripts (even without captions), search channels and playlists, and save results for cross-session recall.

Plan and manage Triplyfy itineraries with interactive maps, places and notes, and search and save flights.

Index source code into a local knowledge base, search with keyword + semantic + hybrid modes.

Project-scoped memory for coding assistants: indexes project code and docs into Qdrant with tree-sitter chunking, LSP integration and hybrid search.

Yandex 360 Tracker, Wiki and Forms: read and write tools with honest read-only annotations.

Local AI memory: neuroscience-inspired Data→Information→Knowledge in SQLite you own.

Persistent agent memory to remember, recall, forget and share, with semantic vector search and agent, user and org scoping.

Podcast editing service: submit raw audio or a URL and get edited episodes with filler words and noise removed, transcripts, show notes and social clips.

Query your product docs from AI tools. Source-cited answers from your indexed documentation.

Brewing knowledge: BJCP 2021 beer styles, hop varieties, malts, yeasts, water profiles, off-flavour diagnosis, recipe suggestions and food pairings.

Portable personal knowledge graph shared across AI clients, with sources like iMessage, browser history, calendar, shell history, Notion and Obsidian.

MCP server for AI agent read/write access to Obsidian vaults via EVC Team Relay.

Search and contribute to a community-verified knowledge base of engineering issues and fixes, with trust scores and fix verification.

MCP server that adds range-anchored (inline) comments to Google Docs — the one comment operation the Google Docs/Drive APIs cannot do — by driving a real logged-in Docs session with Playwright.

Governed local agent memory in plain files: each entry records who wrote it, when and through which gate, and nothing is kept without user approval.

Local-first memory for AI agents in one SQLite file: hybrid BM25 and on-device embedding search with reranking, abstaining instead of returning wrong memories.

Shared context for dev teams: syncs GitHub, Notion and saved decisions into reusable project context that agents load at the start of each session.

Deterministic reasoning engine for agent context compression: extracts structured facts with logical inference, proof chains and truth maintenance.

Local execution memory for coding agents: injects what worked and known fixes before similar tasks and advises whether to continue, pivot or stop.

MCP server for Obsidian Smart Connections. Semantic search using your vault's embeddings.

Search and read the Sovereign AI Blog on self-hosted AI on NVIDIA DGX Spark, and validate SGLang configs for GB10 hardware.

Taiga project management: issues, user stories, tasks, epics, sprints, comments, attachments and wiki via a compact set of tools.

Search, read and edit a self-hosted Wiki.js 2.x, including assets, comments, users and groups, with full-text grep, overwrite protection and a read-only mode.

Vault-native, accountable memory for Claude Code: Markdown is the source of truth, with no LLM on the Stop path and redaction before any derived store.