Outils

From Microsoft: converts PDF, Word, Excel, PowerPoint, HTML and other files or URLs to Markdown so agents can read and search them. A single convert_to_markdown tool over STDIO or local HTTP.

Gives agents long-term memory across sessions with a local knowledge graph of entities, relations and observations, stored in a local JSONL file you can query and edit. Official MCP reference implementation.

Provides a sequential_thinking tool that guides the model to break complex problems into revisable, branching steps. Good for planning, analysis and tasks that need careful iteration. Official MCP reference server.

Official MCP reference server that fetches web pages and converts them to Markdown for the model, reading long pages in chunks from a given offset. Zero config, but it can reach internal network addresses, so use it with care.

Official PaddleOCR MCP: text recognition and layout parsing for images and PDFs (PP-OCR, PP-StructureV3, PaddleOCR-VL), returning text or Markdown. Runs inference locally, or calls the official API or a self-hosted service.

MCP server built into the adaptive Python scraping framework Scrapling: supports plain HTTP requests, browser rendering and stealth fetching, and uses CSS selectors to pass only the target content to the AI, saving tokens and filtering out prompt injection.

Official OpenBB MCP server exposing the OpenBB Platform's equity, ETF, macroeconomic, crypto and news data to agents; enable only the tool categories you need to save context. Needs uv locally; some data sources need your own API keys. Not investment advice.

Browser automation MCP server open-sourced by ByteDance in the UI-TARS-desktop repo: drives pages through Puppeteer accessibility data with an optional vision mode, running locally over stdio or remotely over SSE or Streamable HTTP.

Trading research agent from HKU's HKUDS whose MCP server offers 70+ tools for quotes, financials, backtesting and research across A-shares, Hong Kong and US stocks and crypto. Needs local Python; HK, US and crypto research need no API key. For research only; not investment advice.

Runs the cognee memory engine as an MCP server whose remember, recall and forget tools write documents and conversations into a knowledge graph and vector store for cross-session retrieval. Start it with Docker or uv first; needs an LLM API key by default, or connect to Cognee Cloud.

Writes conversations and business data into Graphiti's temporal knowledge graph so agents can retrieve long-term memory by entity, fact and time. Open source from Zep; start FalkorDB or Neo4j and the server with Docker first, and configure an LLM API key.

Connects to the Arize Phoenix observability platform so agents can query LLM traces and sessions, manage prompts, browse datasets and read experiment evaluation results. Maintained by Arize; needs a self-hosted or cloud Phoenix instance.

Deep research assistant that runs locally with a built-in MCP server: searches arXiv, PubMed, Wikipedia and more over multiple rounds and writes cited reports. Needs Python 3.12+, uv and an LLM (local Ollama or an API key); web search needs SearXNG or similar. Runs take minutes to half an hour.

Unofficial CLI and MCP that let agents work with NotebookLM (now Gemini Notebook): bulk-import sources, ask questions with citations, and generate audio overviews, quizzes, flashcards, mind maps and slides. Uses undocumented APIs and may break; needs a Google sign-in and uv.

Connects to your local or cloud Zotero library: search by title, full text or semantics, read PDFs and annotations, add items by DOI and organize collections. Needs Python 3.10+; for local mode, enable the local API in Zotero settings first.

Gives agents Exa web search and page content fetching, with an optional multi-step research agent and advanced filtered search. Maintained by Exa; the hosted service works anonymously with limits, and signing in or adding an API key raises them.

Uses Obsidian's community Local REST API plugin so agents can list, read, search, append to and delete notes, and insert content at a heading, block reference or frontmatter field. Install that plugin and configure its API key first.

Stores agent memory as local Markdown notes that both people and AI can read, write, search and link, and you can open them directly in Obsidian. Local-first, open source and free, with a paid cloud sync option.

MCP server from the opengeos GeoAI project: segment buildings, roads and vegetation in satellite imagery from text prompts, detect vehicles and ships, spot changes between images, estimate canopy height and fetch NAIP, Sentinel-2 and Landsat imagery. Install from the repo (Python 3.10+).

Search and download arXiv papers and read the authors' LaTeX source section by section, export BibTeX and follow topic updates, with papers cached locally. No API key needed.

Converts PDF, Word, Excel, PowerPoint, images, audio, web pages and YouTube transcripts into Markdown that agents can read and cite. Built on Microsoft MarkItDown; the Docker image lacks audio transcription and image recognition, so build locally for full features.

Unified search across 20+ academic sources including arXiv, PubMed, bioRxiv, Semantic Scholar and OpenAlex, returning deduplicated, normalized results and downloading full text via open-access channels first. Most sources need no API key.

Searches Korean statutes, case law, administrative rules and local ordinances via the Korean Ministry of Government Legislation's Open API, verifies whether cited provisions and cases exist, and compares versions over time. Needs a free API key and Node.js; have a lawyer review results.

Perplexity's official MCP with four tools: web search, grounded Q&A, deep research and reasoning, available remotely or run locally. Requires a Perplexity API key.