A CLI-first skill for AI agents to export documentation subtrees and inspect audit artifacts like manifest.jsonl and report.html.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install docsiphon-doc-corpus-operator
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install docsiphon-doc-corpus-operator using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The docsiphon Doc Corpus Operator is a technical skill designed for AI agents to execute documentation scraping and corpus generation via a CLI-first path. It enables the creation of high-quality documentation subtrees from web sources, providing a host-native secondary lane for agents to acquire context without relying on hosted browser workflows. Using this Openclaw Skills extension, developers can transform disparate web docs into structured datasets suitable for LLM ingestion.
This skill emphasizes a local execution model, utilizing tools like uvx to run exports and generate a comprehensive suite of audit artifacts. It focuses on transparency and local verification, ensuring that the agent can inspect the precise output of a scoped export before moving to larger site scopes or custom profiles.
To get started with this skill in your Openclaw Skills environment, follow these installation steps:
# Install docsiphon via uvx or from the local repo checkout
uvx docsiphon --help
# Run a small scoped export as a first-success test
docsiphon export https://example.com/docs --output ./export-test
Refer to references/INSTALL.md for specific environment configurations and references/DEMO.md for the initial first-success flow.
The skill produces a structured output directory containing various audit and content files. The following table describes the primary artifacts:
| Artifact | Description |
|---|---|
manifest.jsonl |
Line-delimited JSON containing metadata for all exported pages. |
report.json |
A machine-readable summary of the export process and success metrics. |
report.html |
A human-readable visual audit of the documentation export. |
toc.md |
A generated Table of Contents for the exported corpus. |
llms.txt |
A specialized text file optimized for LLM consumption. |
Loading
An automated auditing engine that identifies performance bottlenecks, React 18+ pattern violations, and bundle size issues in frontend codebases.

An AI-driven video editing skill that transforms raw footage into polished 1080p MP4 videos with music and transitions in under two minutes.

A cloud-powered AI video editing tool that automates trimming, polishing, and rendering raw footage into professional 1080p clips.

A powerful skill for automating the reading, cleaning, and writing of CSV data using the pandas library.

A conversational AI video editor that transforms raw footage into polished 1080p MP4 clips through simple text instructions.

An AI-powered video transformation tool that converts long-form horizontal content into platform-ready vertical shorts with captions and smart cuts.








































