microsoft/markitdown A MCP Server Official Active
MCP tool access to MarkItDown -- a library that converts many file formats (local or remote) to Markdown for LLM consumption.
Browse all Model Context Protocol servers with quality scores, stars, languages and maintenance activity.
MCP tool access to MarkItDown -- a library that converts many file formats (local or remote) to Markdown for LLM consumption.
Remote Storage Access: SFTP, S3, FTP, SMB, NFS, WebDAV, GIT, FTPS, gcloud, azure blob, sharepoint, etc.
Share code context with LLMs via MCP or clipboard
AI-native directory visualization with semantic analysis, ultra-compressed formats for AI consumption, and 10x token reduction. Supports quantum-semantic mode with intelligent file categorization.
Filesystem server for legacy, non-UTF-8 files. Auto-detects and converts 24 encodings (CP1251, CP1252, ISO-8859-x, KOI8-R/U, GBK/GB18030), keeps each file's original encoding on write, and manages BOM and line endings. Encoding-aware grep, partial reads with offset/limit for large files, and allowed-directory confinement. Single Go binary, no runtime dependencies.
Production-ready MCP server for intelligent handling of large files with smart chunking, navigation, streaming capabilities, regex search, and built-in LRU caching.
A filesystem allowing for browsing and editing files implemented in Java using Quarkus. Available as jar or native image.
Advanced filesystem operations with large file handling capabilities and Claude-optimized features. Provides fast file reading/writing, sequential reading for large files, directory operations, file search, and streaming writes with backup & recovery.
Surgical editing of existing Word .docx files with formatting preservation, tracked changes, comments, footnotes, and document comparison.
Coherence guard for shared files: when two agents write the same file, the stale writer is denied instead of silently overwriting, then reacquires and retries. Single-host, TLA+-verified. Install: `uvx --from "agent-coherence[mcp]" stale-write-guard-fs`.
Encrypted S3-compatible object storage with continuous cryptographic possession proofs. In-conversation signup (12 GB free), put/get/ls/rm/usage, optional client-side encryption; credentials stay on your machine.
File merger tool, suitable for AI chat length limits.
Windows-native full-disk search over file names, document contents, and OCR'd screenshot text, exposed to AI agents via a read-only local MCP server.
Free file conversion between 690+ formats. Tools: `convert_file` (URL or base64 in → signed download URL out) and `list_conversions`. Covers image, video, audio, document, data, font, ebook, and archive formats. No auth or signup required; remote streamable-HTTP endpoint available (see README).
Rust-powered PDF toolkit: create, read, and analyze PDFs; extract text and entities for RAG; convert to Markdown; split, merge, rotate, and reorder pages; add content, annotations, and form fields; and encrypt documents. Ships in the `oxidize-pdf` PyPI package — run locally with `uvx oxidize-mcp`.
Comprehensive local filesystem MCP server with file management, process control, interactive sessions, async search, and safe delete via Recycle Bin.
Golang implementation for local file system access.
Fast Windows file search using Everything SDK
Read-only MCP server for secure filesystem exploration, searching, and analysis with symlink protection.
Multi-format document MCP server — read, search, OCR, and extract from PDF, TXT, MD, DOCX, CSV, and images. Single Go binary, 12 tools, smart mtime-based caching, directory-locked security.
Read-only directory listing web app with a built-in MCP server: directory browsing, file reading, content search, archive inspection, file hashes, and disk usage tools over a configured root.
Schema-first PDF→JSON MCP for AI agents (0xPdf API). Define a JSON schema, parse invoices/forms (OCR optional), manage schemas, and poll async jobs. `uvx oxpdf-mcp` · [docs](https://0xpdf.io/docs#mcp) · [playground](https://0xpdf.io/samples)
Convert PDF ⇄ Word, Excel, PowerPoint, JPG, HTML, Markdown, EPUB and PDF/A, plus merge, split, compress, watermark, sign, and protect PDFs, and convert images (HEIC, WebP) — 35+ tools from convertica.net. Zero-config via `npx convertica-mcp`.
Deploy files from AI agents to your own FTP/FTPS/SFTP servers — multi-server config, recursive deploy with dry-run and gitignore-like excludes, per-server path jail and read-only mode, FileZilla import, one-command setup for popular MCP clients.
Converts local PDFs into source-linked, card-based standalone HTML readers with notes, highlights, and static publishing.
Token-efficient Python stdio MCP server exposing safe text-file search, reading, and refactoring tools. Tools automatically resolve the file BOM and codepage; edit tools save files with their original encoding and BOM. `uvx text-file-read-and-refactor-mcp`
Ephemeral file transfer for AI agents and CI/CD pipelines — upload any file, get a SHA-256-verified, TTL-enforced link. 4 tools: upload, file_info, list_uploads, delete_upload. Zero-config via `npx -y transfa-mcp`.
The Box MCP server allows third party AI agents to securely and seamlessly access Box content and use tools such as search, asking questions from files and folders, and data extraction.
Hash-anchored file edits for MCP clients. Reads files with line anchors, validates anchors against current content, and rejects stale writes before changing files.
Convert, compress, merge, split, and OCR PDFs, plus convert between 100+ file formats (Word, Excel, PowerPoint, images, ebooks, video) from any AI agent. Backed by the GuruPDF API; free tier with daily credits, zero-config via `npx gurupdf-mcp`.
MCP bridge for [Distill](https://store.juergenkoller.software/en/apps/distill) — AI file renamer that analyzes content (PDF/OCR/Office/email/media) and renames files with descriptive names. Five AI providers (Claude, OpenAI, Gemini, Ollama, Apple Intelligence), pay-per-use credits, folder monitoring, 9 MCP tools.
Model Context Protocol (MCP) Server for reading from Google Drive and editing Google Sheets.
Direct local file system access.
Box integration for listing, reading and searching files
Model Context Protocol server for file system operations. Supports read, write, search, copy, move with security path restrictions.
Access any storage with Apache OpenDAL™
Control your macOS Homebrew setup using natural language via this MCP server. Simply manage your packages, or ask for suggestions, troubleshoot brew issues etc.
Share markdown files and text as clean, readable web links via `plsreadme_share_file` and `plsreadme_share_text`; zero setup (`npx -y plsreadme-mcp` or `https://plsreadme.com/mcp`).
Reduce context usage on file use. Send only unified diffs instead of full files (up to 20-100× fewer tokens), and read large files with targeted `grep`/`sed` instead of full reads (up to 500×). Kernel-enforced path confinement hard-locks the agent to a configured root: no accidental reads or writes outside scope. Standalone for your file access or embed in any MCP server (Rust, Node.js, Python via
Convert Markdown files to professional PDFs with customizable themes, headers, footers, and styling