# Chiplab

> Chiplab is an MCP platform that lets AI agents simulate and validate firmware on virtual instances of real target chips, without physical hardware.

The MCP endpoint is `https://chiplab.veecle.ai/mcp`.
Any MCP-compatible client (Cursor, OpenCode, Claude Code, etc.) can connect to it directly.

The docs below are available both over plain HTTP at the listed URLs and as MCP resources — agents already connected to the MCP endpoint can read them via resource tools instead of fetching.

## Docs

- [Overview](/mcp/resources/docs/overview.md): what Chiplab is and how it works
- [Getting started](/mcp/resources/docs/getting-started.md): quickstart repo and ready-to-run examples
- [Projects](/mcp/resources/docs/projects.md): what a Chiplab project is and how to work with them
- [Workflow](/mcp/resources/docs/workflow.md): the upload-and-run tool sequences, step by step
- [Testing](/mcp/resources/docs/testing.md): running TypeScript test suites against firmware, and reading their results
- [Test API types](/mcp/resources/schemas/chiplab.d.ts): TypeScript definitions of the `simulation` API available to test suites
- [CLI](/mcp/resources/docs/cli.md): `chiplab-cli`, the command-line client, and how to install and authenticate it
- [CI](/mcp/resources/docs/ci.md): running test suites in GitHub Actions with `chiplab-cli`
- [Supported boards](/mcp/resources/docs/boards.md): the `board` definition, its `model` values, UART capture, and what simulation covers
- [Pricing](/mcp/resources/docs/pricing.md): credits, plans, and enterprise options

## Optional

- [Chiplab docs](https://docs.veecle.ai/llms.txt): full documentation
- [Public repo](https://github.com/veecle/chiplab): quickstart guide, ready-to-run examples across boards and languages, and issue tracker
- [Veecle](https://veecle.ai/llms.txt): the company behind Chiplab