en un clic
agent-tools
agent-tools contient 10 skills collectées depuis edgeimpulse, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Author Edge Impulse custom learning blocks (custom ML blocks). Use when asked to scaffold, modify, test, or push a custom learning block — including the training container Dockerfile, parameters.json for machine-learning blocks, train scripts that read X/Y npy splits and write model artifacts (saved_model.zip, model.onnx, TFLite, model.pkl), and the edge-impulse-blocks CLI init/runner/push workflow.
Work with the Edge Impulse Studio, Ingestion, and Remote Management APIs; CLI tools; Linux runners; SDKs; and custom blocks. Use for project and sample management, data upload, model training and testing, deployment builds, device control, job monitoring, automation, or any programmatic Edge Impulse platform integration.
Adapt applications for Arduino UNO Q and Arduino App Lab as apps or reusable bricks, integrate Edge Impulse EIM models, connect the Linux MPU and STM32 MCU through RouterBridge, and build Flask web UIs. Use for UNO Q development, App Lab manifests and CLI workflows, model deployment, camera apps, or porting existing projects into App Lab.
Write C++ application code that uses a downloaded Edge Impulse C++ library, covering both the portable "C++ library" and the hardware-accelerated "C++ library (Linux)" Studio deployment options. Use for desktop or Linux applications, CMake integration, or porting the library into a generic MCU or custom C++ build system. For Arduino, STM32CubeIDE, or Zephyr targets, prefer the dedicated firmware-arduino, firmware-stm32, and firmware-zephyr skills.
Run Edge Impulse EIM models on Raspberry Pi and other Linux systems with the edge_impulse_linux Python SDK. Use when building Python inference loops for custom sensors, images, video, USB microphones, audio classification, keyword spotting, or when debugging ImpulseRunner and AudioImpulseRunner applications.
Integrate an exported Edge Impulse C++ library into STM32CubeIDE projects, including STM32H7 and FreeRTOS applications, sensor sampling, ISR-safe buffering, classifier tasks, linker and memory configuration, and STM32 SDK porting. Use when writing or debugging Edge Impulse inference firmware for STM32 microcontrollers.
Integrate Edge Impulse Zephyr Module deployments into Zephyr or Nordic nRF Connect SDK applications, including west manifests, CMake, Kconfig, devicetree sensors, static-buffer inference, clean builds, flashing, and porting across boards. Use for nRF5340 and other Zephyr-supported targets.
Write Arduino application code that uses a downloaded Edge Impulse library. Use when asked to write a sketch, integrate an Edge Impulse model into Arduino code, or run inference on an Arduino board.
Author Edge Impulse custom deployment blocks (Enterprise). Use when asked to scaffold, modify, test, or push a custom deployment block — including Dockerfile, parameters.json schema, the build script that reads deployment-metadata.json and produces deploy.zip, and the edge-impulse-blocks CLI workflow.
Interact with the arduino-router daemon on Arduino UNO Q using MessagePack RPC. Use when creating custom C++, Python, or other clients for Linux MPU to STM32 MCU communication without the Arduino App Lab Python Bridge API.