#compression

Ziplark's MCP server (ziplark-mcp) lets an AI agent list, extract, test and create archives — ZIP, 7z, tar and gz/bz2/xz/zst/lz4, plus read-only RAR/RAR5 and ISO — all with AES-256 and a single zip-slip guard on every extraction path. Read tools are always on; writes are gated behind --allow-write. Part of Ziplark, one small Rust engine that also ships a desktop app and a CLI. Free and open source (MIT), cross-platform (macOS, Windows, Linux).

Agentic AI instruction encoding. 86.8% smaller than JSON. Inference-free decode by table lookup. 342 opcodes, 26 namespaces, three conformant SDKs. Confirmed over LoRa mesh radio. Apache 2.0 with express patent grant.

Compress images up to 90% smaller. 5 tools: compress_image, image_info, compress_folder, compress_document (PDF/DOCX/PPTX — unique!), convert_format. Supports JPEG, PNG, WEBP, AVIF, HEIC, RAW. Free API key at deflato.com.

AI-powered code analysis and context optimization tools. Slim large code files to relevant sections (save 60-80% tokens), compress text, analyze conversations for noise, and search past chat knowledge bases. Supports x402 micropayments on Base.

Compresses large blocks of text, code, or conversation history into hyper-dense Reasoning Maps optimized for LLM context windows. Saves agents 80%+ on token costs. Stateless. Pay $0.001 USDC per compression via x402 on Base network.

NanoImage MCP gives Claude the ability to process images directly in the conversation. Supports 9 tools: compress, resize, crop, rotate, flip, convert (JPG/PNG/WebP), blur, grayscale, and get image metadata. Images are processed server-side via Sharp.js and never stored. Free, public, no authentication required. Rate limit: 20 requests/minute/IP.

A Model Context Protocol (MCP) server for comprehensive image editing operations including resizing, format conversion, cropping, compression, and more.

Mirror of

An MCP server for FFmpeg

An MCP tool that provides AI with the ability to compress and decompress local files.