statespace
Interactive web apps for AI agents
claude mcp add --transport stdio statespace-tech-statespace docker run -i statespace/statespace \ --env PORT="8000" \ --env STATESPACE_TOKEN="<optional>"
How to use
Statespace is a declarative framework that lets you build interactive web apps for AI agents using Markdown. The server hosts your Markdown-based apps and exposes tools, components, and workflows that agents can interact with over HTTP. It emphasizes simplicity by keeping content in Markdown and rendering dynamic components and workflows through a Rust binary exposed via a Docker image. To use it, run the provided container, point agents or clients at the served URL, and let them execute the available tools and components within your Markdown docs. The platform emphasizes token-based access control and easy deployment to the cloud or locally.
How to install
Prerequisites:
- Docker or an alternative container runtime
- Access to pull the Statespace Docker image (statespace/statespace)
Installation steps:
-
Install Docker: follow instructions at https://docs.docker.com/get-d started/
-
Pull and run the Statespace image:
docker run -i statespace/statespace
-
Access the server:
- By default the service exposes port 8000 inside the container. Map it to your host if needed, for example:
docker run -p 8000:8000 -i statespace/statespace
-
Verify startup by visiting http://localhost:8000 in your browser or by querying the API endpoints used by your agents.
Optional: set environment variables if you need authentication or custom port configuration, for example:
docker run -p 8000:8000 -e PORT=8000 -e STATESPACE_TOKEN=your_token -i statespace/statespace
Additional notes
Tips and common considerations:
- Default port is 8000 inside the container; expose and map it with -p as needed.
- If you enable token-based authentication, set STATESPACE_TOKEN in the container environment and pass the token with your agent requests.
- The Statespace image is designed to run Markdown-driven apps; ensure your app content follows the expected Markdown and YAML configurations for Tools, Components, and Instructions.
- For local development, you can bind-mount your app directory into the container if the image supports it, or build a custom image with your app pre-baked.
- If you encounter networking issues, verify that the mapped port is accessible from your client and that your firewall allows traffic to the mapped port.
- Check the official documentation for the latest image tag or deployment instructions and for any required runtime options.
Related MCP Servers
excelize
Go language library for reading and writing Microsoft Excel™ (XLAM / XLSM / XLSX / XLTM / XLTX) spreadsheets
anyquery
Query anything (GitHub, Notion, +40 more) with SQL and let LLMs (ChatGPT, Claude) connect to using MCP
headroom
The Context Optimization Layer for LLM Applications
Wax
Sub-Millisecond RAG on Apple Silicon. No Server. No API. One File. Pure Swift
pgmcp
An MCP server to query any Postgres database in natural language.
decipher-research-agent
Turn topics, links, and files into AI-generated research notebooks — summarize, explore, and ask anything.