Embedded-AI-Harness

View on GitHub

AI Closed-Loop Programming for embedded systems — the AI writes, flashes, and tests firmware on real hardware until the tests run clean. Spec to silicon, hands off.

★ 18758 forksPythonMITSensorsIot

A Claude Code-oriented harness for developing and testing ESP32 firmware in a closed loop on real hardware. A Raspberry Pi testbench exposes board flashing, debugging, and hardware tests through HTTP and MCP tools so the AI can use test failures to guide fixes.

Use Cases

Write and iteratively fix ESP32 firmware using test feedbackFlash firmware to real boards remotelyRun hardware-in-the-loop tests for WiFi, BLE, MQTT, and RFDebug and observe firmware on Raspberry Pi test benchesAutomate embedded product development from specification to verified firmware

Built With

Language: Python
Frameworks: Claude Code · MCP · ESP-IDF · PlatformIO · pytest

Embedded-AI-Harness

Use Cases

Built With

Tags