AI Engineering

GPIO Bare-Metal configures MCU GPIO without HAL for input/output modes, alternate functions, pull resistors, and pin-change interrupts. Use it for LED, button, and peripheral pin setup on STM32, nRF52, or ESP32-class boards.

ML Iteration generates grounded, ranked next steps when an ML experiment is stuck or underperforming. Use it after initial runs to compare alternatives, review what was tried, and decide what to change next.

ML Planning turns an ML goal into a grounded, step-by-step implementation plan. Use it when you need an architecture, pipeline, or build plan for a machine learning project.

ML Research helps compare ML and AI approaches using verified framework documentation and model references. Use it when you need to understand how a method works or survey framework capabilities.

MMIO and Bit Manipulation guides safe memory-mapped I/O for bare-metal firmware. Use it when reading or writing hardware registers with volatile access, masks, read-modify-write patterns, alignment, or endianness concerns.

OPC routes tasks through a digraph of specialist agents for building, reviewing, analyzing, and brainstorming. It uses independent evaluation gates so the agent that does the work does not evaluate it.

Peripherals From Datasheet guides writing MCU peripheral drivers from reference manuals by reading register maps, init sequences, and timing notes. Use it when porting or debugging register-level drivers without a vendor HAL.

Using Leeroopedia establishes the lookup-first workflow for ML and AI conversations. It tells the assistant to use Leeroopedia or web sources, cite claims, and check for config or API mistakes before answering.

Add Compact Command adds a /compact session command to manually compact conversation history in long-running Claude sessions. Use it to reduce context rot while keeping the same logical session and honoring main-group or trusted-sender access rules.

Add Gmail integrates Gmail into ClaudeClaw as a tool or full channel. It is used to read, send, search, draft, and optionally trigger the agent from incoming email after OAuth setup.

Claude Code Rescue delegates substantial investigation, implementation, or follow-up work to Claude Code through the tracked-job runtime. Use it when Claude should diagnose, edit, test, or continue a task rather than just report review findings.

ClaudeClaw Customization asks clarifying questions and then edits the code to add channels, integrations, triggers, or other behavior changes. Use it when you want to modify how ClaudeClaw works or migrate from plugin mode to developer mode.