Open-Source-Legal/OpenContracts

The open document intelligence platform for builders and hackers - DMS for the agentic world

What it solves

OpenContracts provides a structured substrate for document intelligence, moving beyond simple file storage to create a programmable citation graph. It solves the problem of "leaf" documents (like PDFs) that lack connections to other documents, statutes, or precedents, which often leads AI agents to hallucinate edges or fail when encountering unresolved references.

How it works

The system transforms a repository of documents into a navigable graph where documents are nodes and citations are edges. It uses a a modular pipeline for parsing, embedding, and thumbnailing documents.

Key components include:

  • Citation Graph: Automatically detects and resolves statutory citations to create a web of references.
  • AI Agents: Python-based agents that can reason over the curated citation graph and provide grounded, cited answers.
  • MCP Server: A Model Context Protocol implementation that allows external AI tools (like Claude or Cursor) to search the corpus and walk citation edges without custom glue code.
  • Structured Extraction: A "fieldset" system that uses LLM-powered queries to extract specific data across hundreds of documents into a grid for human review.
  • Human Annotation: A layer where humans define label schemas and annotate precise text spans on PDFs, providing the ground truth that AI agents build upon.
  • APIs: A GraphQL and REST API that exposes the entire graph for external applications.

Who it’s for

It is designed for teams working with large-scale document repositories, such as legal archives, research libraries, engineering knowledge bases, or internal policy folders, who need tot turn their documents into a machine-readable and human-navigable knowledge graph.

Highlights

  • Model Context Protocol (MCP): Native support for bringing your own agent via a standardized server.
  • Human-in-the-loop: Treats human annotation as the ground truth for the citation graph.
  • Layout-faithful PDFs: Precise text-to-coordinate mapping for annotations on PDFs.
  • Version-controlled Corpuses: Git-like branching and forking for document collections and their annotations.
  • Pluggable Pipeline: Swappable parsers and embedders for different file formats.

Related

  • Project
    vercel-labs/lat.md

    lat.md is an npm‑distributed CLI that creates a markdown‑based knowledge graph for a codebase. It lets you write linked documentation in `lat.md/`, annotate source files with `@lat` comments, and validates the graph. Agents and developers can search it semantically (offline MiniLM or remote OpenAI/Vercel embeddings) and use commands like `lat check`, `lat locate`, and `lat expand` to keep design decisions, constraints, and test specs in sync with the code.

  • Project
    gszhangwei/open-spdd

    A methodology and CLI tool that transforms AI coding prompts into structured design contracts using the REASONS Canvas framework to ensure precise, traceable code generation.

  • Project
    pipeshub-ai/pipeshub-ai

    PipesHub is an open‑source, self‑hostable platform that connects LLMs to enterprise data (Slack, Google Drive, GitHub, etc.). It builds a permission‑aware knowledge graph, provides citation‑backed answers, supports any LLM via LangChain, offers 50+ connectors, multimodal document handling, a no‑code agent builder, and SDKs for Python, TypeScript, and Go. Deployable with a single Docker‑Compose installer.

  • Project
    invertase/docs.page

    An open-source platform that transforms public GitHub Markdown files into modern, agent-ready documentation sites with built-in AI chat and MCP support.

  • Project
    federicodeponte/opendraft

    An open-source AI research engine that generates long-form, source-grounded academic drafts with citations verified against CrossRef, OpenAlex, and Semantic Scholar.