Skip to main content
This page describes what your AI agent can do with Chiplab. You don’t need to call anything yourself, just ask your agent to run or test your firmware and it takes care of the rest. This is a reference for the curious, not a set of instructions for you to follow.
Chiplab can execute your firmware on a chip-accurate virtual board and return the console output, no physical hardware required. Chiplab runs your firmware on a virtual instance of the target board via a simulation platform (today that means Renode, with QEMU and more planned). Every supported board runs the same binary, the same peripherals, and the same interrupt timing as its physical counterpart, so what you see in Chiplab reflects what you’d see on real hardware. Depending on what’s live when you connect, this might be a single call or a small sequence (e.g. issuing an upload slot, uploading the artifact, then triggering the run). Your agent’s MCP client discovers the exact tool calls and parameters live; ask Chiplab if you want to see them yourself.

​
Parameters

​
string
required
The identifier of the firmware binary to simulate. Your agent obtains this automatically before running.
​
string
required
The slug identifying the virtual board to emulate. See Supported Hardware for the full list. Example values include stm32f4_discovery and nrf52840_dk.

​
Response

​
string
A unique identifier for this simulation run.
​
string
The captured UART/console output produced by your firmware during the simulation.
​
string
Simulator diagnostics emitted during the run (e.g. faults or simulation-level errors), separate from your firmware’s own UART output.

​
Example calls

Run a firmware binary on a target board:
A successful response looks like this:
Not sure which board fits your firmware? Ask Chiplab about individual board specs; board recommendation tooling is on the way.

​
Chip accuracy

Chiplab’s virtual boards are not generic CPU emulations. Each board runs your exact firmware binary against a faithful model of the target chip’s peripherals, memory map, and interrupt controller, so results are directly comparable to real hardware outcomes.

​
Knowledge corpus

Every run deposits observed behavior into Chiplab’s knowledge corpus, indexed by chip family, failure pattern, and board configuration. Chiplab never collects your firmware code or binaries as part of this, only the chip-level behavior observed during the run.