MCP Servers tools for Local / Ollama

🔥 MaxKB is an open-source platform for building enterprise-grade agents. 强大易用的开源企业级智能体平台。

Use any LLMs (Large Language Models) for Deep Research. Support SSE API and MCP server.

The open-source context layer for AI agents. PipesHub turns your company's knowledge (Slack, Drive, Jira, GitHub, Microsoft 365 and 40+ connectors) into a permission-aware workspace that agents can search, grep, navigate and cite. MCP, SDKs and built-in agents. Self-hosted.

Personal AI agent harness for macOS, Windows, Linux, Android & iOS. New: Skales Scout, lasting AI agents that keep working on your desk, each optionally in its own virtual machine. Chat, coding (Skales Code), Flow, automations, computer use, browser, voice, MCP, local LLMs (Ollama). BYOK, 26 providers. No Docker, no terminal.

Jev-compatible API server for open decision models. Ollama for decision models: pull and serve 29 model families (Laya, pplx-decider, Clef, Winnow, coba...) locally or self-hosted behind /v1/systemone. The TypeSafe SDK works unchanged: set its base URL and pick an open model. CLI, desktop app, Docker, MCP server, public bench.

Local RAG (on-premises) with MCP server

Provides multi-cluster Kubernetes management and operations using MCP, featuring a management interface, logging, and nearly 50 built-in tools covering common DevOps and development scenarios. Supports both standard and CRD resources

Ask questions across your Markdown notes using a fully local Graph RAG engine. Built for Obsidian vaults, works with any folder of Markdown files. Extracts entity-relation triples from wikilinks & YAML frontmatter, retrieves answers via hybrid search (vector + BM25 + temporal). Multilingual. No cloud. Runs on Ollama.

Assistant vocal local en francais : Claude ou Ollama (offline), domotique Hue, OBS, agenda, navigateur, appels Twilio, serveur MCP. Python.

本机优先的 AI 办公工作台 · A local-first AI office workbench. Turn requests and source material into editable PPTX, DOCX, XLSX and HTML files. Choose your own models, Skills and MCP tools; inspect execution and results.

Self-hosted AI studio exposing local image, video and music-video generation, upscaling, RAG and code intelligence, with renders queued on your own GPU

by Yali Pollak - A local-first router that sits under Claude Code (and Codex/Gemini CLI) and sends each prompt to the cheapest capable model, with three-layer token compression and automatic provider fallback — protecting your premium quota and cutting cost with zero config on a Claude subscription. Credibly engineered: 1,900+ tests, an independent RouterArena benchmark placement, and an honest lo

Self-hosted mem0 for Claude Code: Qdrant vector search, Neo4j knowledge graph, Ollama embeddings and session hooks for automatic cross-session memory

Queryable second brain over your scattered notes, docs, PDFs and chat exports: hybrid retrieval (vector + BM25 fused with RRF), section-level citations with an optional claim-by-claim verification pass, and an Obsidian-style wikilink graph — all local, any OpenAI-compatible API or fully offline via Ollama. pip install loci-rag then run loci-mcp

Automate cybersecurity assessment of documents, questionnaires and reports to surface risks, compliance gaps and remediations, using a RAG knowledge base

Query multiple Ollama models and combine their responses for diverse perspectives on a single question

Multi-round brainstorming debates between LLMs (GPT, Gemini, DeepSeek, Groq, Ollama and more) to explore ideas from diverse perspectives

Single-binary local manager and search engine that stands up a llama-server fleet sized to your GPUs by gguf-parser for VRAM-aware multi-GPU placement

Self-hosted server suite for Odoo ERP 15-19, plus GitHub, Portainer, OCA and Teams, with a browser terminal, Qdrant/Ollama memory and Bulgarian localization

Ask what went wrong with your own Ollama calls: lost prompt cache and the token where the prompt stopped matching, context overflow, a model running partly on CPU, tool schemas crowding the prompt. Last turn in numbers; which local models fit in video memory. Read-only

Self-hosted digests of your Telegram channels over MCP: AI summaries, the last digest and raw per-channel messages, with OpenAI, Anthropic or Ollama

Local-first developer knowledge graph: watches project files, extracts entities and relationships with LLMs, and offers queries, a web dashboard and a CLI

#️⃣ 🏠 🍎 🪟 🐧 - Self-hosted knowledge backend for agents with hybrid vector and keyword search and container-isolated indexes

Retrieve memory from across your Claude Code sessions on demand

Local Clef-Flash decision model for coding agents: clef_decide returns probability distributions over your options (choice/score/yes-no) via llama.cpp, fully offline, one-command install

Judgment calls on small, cheap models: yes/no, pick-one and rate questions about any text, answered YES, NO or UNSURE. Runs on TypeSafe's Jev, open models on Ollaya, Ollama or a local NLI model; each judges up to 1,000 texts per call

Self-hosted Markdown knowledge base that doubles as persistent memory for AI agents: a notes app you browse and edit (wikilinks, backlinks, graph), 18 MCP tools over Streamable HTTP with OAuth (works as a claude.ai connector) or stdio. Hybrid search (Postgres full-text + pgvector, RRF, optional reranker) with section-level reads and conflict-safe appends. npx -y kybase-mcp

Hybrid computer-use runtime combining the accessibility tree, Chrome DevTools Protocol and vision into structured context, with offline support via Ollama

Local-first agent memory and knowledge base: hybrid dense and BM25 retrieval, document and webpage ingestion and optional RAG answers via local Ollama

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

German public procurement data (OCDS): semantic search, tender matching against company profiles and structured filtering

Local temporal knowledge graph with six read-only retrieval tools, Ollama inference, embedded LadybugDB, and a separate human-approved CLI write queue

AI execution pipeline with NAEF logic constraints, VERITAS build gates, semantic vault integration and Tri-Node Cortex validations to curb hallucination

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

Bridge to a local Ollama server for running Llama, Mistral, Qwen and other local models

Chat, generate, pull, list and inspect models with Ollama

Query SQLite, PostgreSQL, BigQuery and Databricks in natural language through a business semantic layer: glossary, metrics and a data dictionary

Rename files with descriptive names via Distill, which analyzes PDF, OCR, Office, email and media content, with folder monitoring

Compresses AI chat sessions into a compact typed knowledge graph, with optional offline extraction via Ollama

Web design analysis: layout extraction, motion detection, quality scoring and semantic search, turning web pages into structured design knowledge

Local meeting pipeline: Whisper transcription plus Ollama summaries, action items and decisions. Reference implementation of the memnex spec

Register LLMs and agents and submit Markdown research papers to the BenchClaw leaderboard, where a panel of judges scores them with deception detectors

Live air quality, pollen and weather data from Ambee

Read-only health checks for a local vLLM instance: liveness and real-completion tiers, GPU and VRAM status, service status and serve-flag introspection

Live web grounding for local and cloud LLMs: web search, page fetch, token-capped cited evidence packs and an agentic search-and-read research loop

Bridge to a locally running Ollama or vLLM instance for pure language processing (writing, summarizing, classifying, structured extraction), without giving the calling agent tools or file access. pip install mcp-ollama-vllm