About inhalt
Inhalt is an open-source headless CMS for managing content through AI tools and MCP clients. Typed schemas constrain edits to defined fields, while drafts, review, publishing, and a separate read API support content delivery.
Key features
- 01Read, query, and patch content through MCP
- 02Validate edits against typed content schemas
- 03Preview drafts and publish after review
- 04Provide a headless read API
- 05Scope access by entries and actions
- 06Self-host under the MIT license
Use cases
- 01Edit CMS content through an MCP client
- 02Review draft changes before publishing
- 03Restrict clients to specific entries and actions
MCP support
Inhalt exposes its content tree as typed tools to MCP clients. Clients can inspect schemas, locate entries, and submit validated field changes.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEInhalt
The CMS your AI tools can run.
Inhalt is an open-source, MCP-native CMS. Instead of clicking through a traditional admin UI, the content layer is exposed through the Model Context Protocol (MCP). Any MCP client connects to the Inhalt server and you create and edit content just by chatting.
The protocol is the API. The client reads your content schema, finds the right entry, and proposes a clean, typed patch against real fields, never freeform text.
- Website: inhalt.tech
- MCP endpoint:
https://inhalt.tech - License: MIT
- Self-hostable: one container, your database, your keys. A hosted edge is also available.
How it works
Three steps.
- Connect the endpoint. Add one block to your MCP config. Inhalt exposes your whole content tree as typed tools.
- Operate on content. The client reads the schema, finds the right entry, and proposes a clean patch against real, typed fields.
- Save and publish. Every change is checked against your schema, saved, and pushed to your live read API in one step.
See docs/getting-started.md for the MCP config and a full walkthrough.
Features
| Feature | Description |
|---|---|
| MCP-native | The protocol is the API. Read, query, and patch content through tools any client already understands. |
| Typed schemas | Model content once. Clients work strictly inside your fields, validated and never freeform. |
| Drafts and review | Stage edits, preview them, and publish only when they are ready. |
| Headless delivery | Query from anywhere over a clean, fast read API. |
| Scoped access | Grant a client exactly the entries and actions it should touch. |
| Self-host | One container, your database, your keys. MIT licensed, end to end. |
Supported clients
Inhalt works with every MCP client, including Claude, Cursor, VS Code, Zed, Windsurf, and Cline.
Documentation
- Getting started - connect an MCP client and operate on content.
- Concepts - projects, content types, entries, and the draft/publish model.
- API reference - the Read API and MCP API.
- Development - run the repo locally, environment, and scripts.
- Code organization - project layout and where new code goes.
- Copy & tone - writing guidelines for product copy.
- Roadmap - planned work (auth, self-host).
License
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.