Skip to content

archivebox-api

A Pythonic ArchiveBox API wrapper, MCP Server, and A2A Agent for the agent-utilities ecosystem — programmatic access to an ArchiveBox web-archiving instance for agentic AI.

Official documentation

This site is the canonical reference for archivebox-api, maintained alongside every release.

PyPI MCP Server License GitHub

Overview

archivebox-api wraps the ArchiveBox REST and CLI surface with typed, deterministic MCP tools, and ships a Pydantic-AI agent for conversational operation. It provides:

  • Api — a requests-based REST client over the ArchiveBox API, organized by capability (authentication, core catalog, CLI), built on agent-utilities.
  • Action-routed MCP tools — three consolidated, togglable tool modules (archivebox_authentication, archivebox_core, archivebox_cli) that minimize token overhead in an LLM context.
  • An integrated graph agent — a Pydantic-AI agent server (console script archivebox-agent) exposing the same capability over A2A and the Agent Web UI.

Explore the documentation

  • Installation — pip, source, extras, and the prebuilt Docker image.
  • Deployment — run the MCP and agent servers, Docker Compose, Caddy + Technitium.
  • Usage — the MCP tools, the Api client, and the CLI.
  • Backing Platform — deploy ArchiveBox with Docker.
  • Overview — ecosystem role, configuration, and architecture.
  • Concepts — the CONCEPT:ABOX-* registry.

Quick start

pip install "archivebox-api[mcp]"
archivebox-mcp                   # stdio MCP server (default transport)

Connect it to an ArchiveBox instance:

export ARCHIVEBOX_BASE_URL=http://your-archivebox:8000
export ARCHIVEBOX_USERNAME=admin
export ARCHIVEBOX_PASSWORD=your-password
archivebox-mcp --transport streamable-http --host 0.0.0.0 --port 8000

See Installation and Deployment for the full matrix (PyPI extras, Docker image, all transports, the agent server, reverse proxy, DNS).