App Inventor MCP
Enables AI agents to build and edit MIT App Inventor projects as .aia files, supporting screens, components, properties, blocks, assets, and project merging.
README
App Inventor MCP
An MCP (Model Context Protocol) server that lets AI agents build and edit MIT App Inventor projects. The agent creates real .aia files on disk, which you then import into App Inventor via Projects > Import (.aia).
Part of the AIA Stitch family. Sister project: aia-stitch (web UI for merging projects).
Why
App Inventor is a visual drag-and-drop tool, which historically locked out AI agents. But an .aia file is just a zip: screens are JSON documents (*.scm), blocks are XML (*.bky), assets are plain files. This server exposes those facts as tools any MCP client can call, so an agent can scaffold an entire multi-screen app for you before you ever open the designer.
Setup
Requires Node 18+.
npm install
node server.mjs
Add to your MCP client config (Claude Desktop, opencode, etc.):
{
"mcpServers": {
"app-inventor": {
"command": "node",
"args": ["/absolute/path/to/app-inventor-mcp/server.mjs"]
}
}
}
Tools
| Tool | What it does |
|---|---|
aia_create_project |
New .aia with a Screen1, valid properties |
aia_add_screen |
Add a screen |
aia_inspect |
List screens, components, assets, properties |
aia_get_screen |
Read one screen's components, properties and blocks XML |
aia_add_component |
Add a component, optionally nested in a parent arrangement |
aia_set_property |
Set a property on the screen or a component |
aia_set_blocks |
Replace a screen's blocks XML |
aia_add_event_handlers |
Generate event handlers (when Click do...) without knowing Blockly XML |
aia_add_asset |
Embed a local file (image, sound) into the project |
aia_merge |
Merge multiple .aia files into one |
Testing
node test.mjs # library roundtrip: create, author, merge
node smoke.mjs # full MCP protocol handshake + tool calls over stdio
node stress.mjs # scale (25 screens / 120 components), concurrency, abuse, recovery
Honest limits
Components and screen properties are fully supported. Blocks authoring works at two levels: generated event handlers via aia_add_event_handlers (property setters and screen navigation), or raw Blockly XML via aia_get_screen / aia_set_blocks. There is no high-level vocabulary for every block type yet. UI-first apps work end to end today; exotic logic may want finishing touches inside App Inventor.
All file operations are serialized per path inside the process, so pipelined/concurrent calls from an agent are safe. Cross-process access to the same .aia at the same time is not guarded; don't edit one file from two servers at once.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
E2B
Using MCP to run code via e2b.