What problem does it solve? When OpenClaw behaves differently across local tests, live models, providers, code mode, Tool Search, Crabbox, or CI, it is hard to know which boundary is failing. This Skill replaces guessing with a structured debugging loop that picks the right logs, probes, and proof path before any code is changed. ## Core Features & Use Cases - Boundary-first debugging loop: Identify whether the fault lies in config, tool construction, provider payload, fetch, SSE streaming, transcript replay, worker/runtime, packaging, or CI before patching. - Targeted transport logging: Use environment flags like OPENCLAW_DEBUG_MODEL_TRANSPORT, OPENCLAW_DEBUG_MODEL_PAYLOAD, OPENCLAW_DEBUG_SSE, and OPENCLAW_DEBUG_CODE_MODE to capture request shapes, tool names, and stream timing without global debug noise. - Proof selection guidance: Match each fix to the narrowest verification, from targeted Vitest runs and pnpm build for worker/dist changes to Crabbox for Docker, package, Linux, and CI-parity proof. - Use Case: A model call hangs with no output. Enable SSE event logging, confirm the fetch response arrived but the first SSE event never did, and isolate the issue to the provider stream rather than tool execution. ## Quick Start Ask the assistant to use the openclaw-debugging skill to identify the failing boundary and enable the right debug flags for the OpenClaw issue you are seeing.
