omcp
omcp is a community-driven open-source hard fork of the Model Context Protocol specification, providing a permissionless, transparently governed standard for connecting AI models to tools and data.
At a Glance
Freely available open-source specification under Apache-2.0 license.
Engagement
Available On
Alternatives
Listed Oct 2026
About omcp
omcp is an open protocol specification for connecting AI models to tools and data, developed and maintained by Enclawed, Inc. It is a community-driven hard fork of the Model Context Protocol (MCP), created to keep the protocol an open, permissionless standard governed transparently on GitHub. The specification is generated automatically from source on every commit and published as both HTML and PDF at the project's GitHub Pages site.
What It Is
omcp is a protocol specification — not a runtime library or SaaS product — that defines how AI models communicate with external tools and data sources. It occupies the same category as the Model Context Protocol it forked from, providing the wire protocol, schema field names, method names, and protocol version identifiers that clients, servers, and SDKs use to interoperate. The schema is defined first in TypeScript and also published as JSON Schema for broader compatibility.
Governance Philosophy
The central motivation for the fork is governance, not technical divergence. The README states the project exists to keep the protocol "an open, permissionless standard: developed asynchronously, reviewed transparently on GitHub, and decided on technical merit rather than on closed-door approval." Contributions are evaluated on five criteria: good coding practices, solving a real and unaddressed problem, a clear problem statement, working unit tests with verifiable data, and a reference documentation chapter. There is no sponsor to find, no working group to join, and no approval to request.
Drop-in Compatibility
omcp is explicitly designed to be drop-in compatible with the broader MCP ecosystem. According to the project README, the wire protocol, schema field names, method names, and protocol version identifiers are unchanged from the upstream MCP specification, meaning existing clients, servers, and SDKs interoperate without modification. The fork is in governance and development process, not in what goes over the wire.
Specification Versioning and Proposals
The project maintains a versioned specification table with clearly labeled statuses. As of the latest build, the current stable version is dated 2026-07-28, with a draft version also available. Earlier versions (2025-11-25, 2025-06-18, 2025-03-26, 2024-11-05) are marked superseded but remain accessible. Design decisions are tracked as SEPs (Specification Enhancement Proposals), with 74 proposals rendered from the seps/ directory. Each proposal corresponds to a pull request, creating a transparent design record.
Update: Version 2026-07-28 (Latest)
The current stable specification is version 2026-07-28, built from commit f6d5b3f on 2026-09-26. The site is rebuilt automatically from docs/specification, schema/, and seps/ on every change to main. A draft version is also published continuously. The project was created in September 2026 and has been actively updated, with the last push recorded on 2026-09-26. The specification and code are licensed under Apache-2.0; documentation (excluding specifications) is under CC-BY-4.0.
Lineage and Attribution
omcp forked from the Model Context Protocol, which was originally created by David Soria Parra and Justin Spahr-Summers. The project explicitly states it is not affiliated with, endorsed by, or sponsored by the MCP project, Anthropic, PBC, or LF Projects, LLC. The omcp name and logo are trademarks of Enclawed, Inc., while the specification itself is open under Apache-2.0.
Community Discussions
Be the first to start a conversation about omcp
Share your experience with omcp, ask questions, or help others learn from your insights.
Pricing
Open Source
Freely available open-source specification under Apache-2.0 license.
- Full protocol specification (HTML and PDF)
- JSON Schema and TypeScript schema
- All versioned releases
- 74 Specification Enhancement Proposals
- Self-hostable build via npm
Capabilities
Key Features
- Open protocol specification for connecting AI models to tools and data
- Community-driven hard fork of the Model Context Protocol
- Drop-in compatible with existing MCP clients, servers, and SDKs
- Specification auto-generated from source on every commit
- Published as HTML and PDF for every version
- TypeScript-first schema also available as JSON Schema
- Versioned specification with stable, draft, and superseded releases
- 74 Specification Enhancement Proposals (SEPs) tracked as pull requests
- Transparent, GitHub-based governance with no closed-door approval
- Apache-2.0 licensed specification and code
