#archive

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).

Read-only Model Context Protocol access to the Zandoria Herald, the national newspaper of the Republic of Zandoria. Connect any MCP-aware client — Claude Desktop, Cursor, custom SDK clients — to a public Streamable-HTTP endpoint and read the same edition citizens read. Five tools cover today's edition, any historical edition by number, full article bodies by slug, keyword search across the archive, and region profiles for the four continental bureaux. Bilingual (English / Esperanto). Stateless, CORS-open, no authentication required. Editions close at 02:00 UTC and are filed in the National Archive.

MCP server for Archivarix Tube Search — search archived YouTube videos including deleted content. Channel discovery, full-text search, video metadata, transcripts, and AI-generated summaries via the Wayback Machine. 1.5B+ videos indexed since 2005.

A MCP server for parsing and analyzing HAR (HTTP Archive) files

Mirror of

A Model Context Protocol (MCP) server for the Internet Archive's Open Library API that enables AI assistants to search for book and author information.

Model Context Protocol (MCP) PRIDE API Server implementation.