Clawddocs Documentation Assistant

Problem

SkillHub documentation is split across providers/, gateway/, concepts/, automation/, and install/. When users ask 'how do I configure X', 'why isn't X working', or 'how do I deploy it', plain search can return noisy results and manual triage requires context switching across setup, troubleshooting, configuration, and automation paths.

How It Works

Clawddocs uses decision-tree navigation to classify the question first, then routes it to the relevant documentation area. It supports document processing, document navigation, and content extraction. Inputs can include file paths, format options, a question, or a knowledge-base scope; outputs can include conversion results, page information, answers, and reference links. Parameters such as content, mode, max_retries, and skip_steps control the run, with mode accepting json, text, or markdown, and skip_steps used for resumable execution. The skill can produce intermediate artifacts and classify failures into configuration, runtime, network, or LLM timeout issues.

Boundaries

It is better suited to structured documentation lookup, format conversion, and content extraction than to complex decisions that require human judgment. If the input format is unclear, context is missing, or the runtime environment is incomplete, results may be unreliable. API Key values should be managed through secure environment variables and kept out of version control.