open-source-legal/opencontracts is an open-source document intelligence platform that covers annotation and extraction for contract and legal documents, along with question answering. It is developed in Python and distributed under the MIT license, and the latest release in the repository is v3.1.1.
The platform analyzes a document repository and builds a citation graph (citation graph) on the basis of human annotations and structured extraction. A citation graph represents the evidentiary links between documents as a graph structure. The resulting graph is served through a GraphQL and REST API, a Model Context Protocol (MCP) server, and a React UI. MCP is the standard interface AI agents use to integrate with external systems.
The README describes the platform as a document management system (DMS) for agents. Because it can be self-hosted, teams can run contract document analysis, evidence linking, and AI extraction workloads on their own servers. The repository was created on October 24, 2022, and has 1518 stars.
That said, the material reviewed here does not summarize the changes in v3.1.1, and it does not include any evaluation data measuring extraction accuracy. Among the scores assigned in the research record, the only filled-in item is momentum 3.2 (on a 10-point scale), while the novelty, technical, and quality items have no values. The structure and the way it is delivered can be confirmed, but how accurately AI extraction actually performs on contract documents is hard to judge from this material alone.