An AI coding agent skill to create, read, and structure documents, tasks, and whiteboards via the unified Craft MCP v2 API.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install craft-skill
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 craft-skill using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
This skill enables an AI coding agent to interact directly with Craft documents through the Craft Model Context Protocol (MCP) v2. It unifies complex note-taking and document management workflows into a streamlined system, making it an essential component of Openclaw Skills. Using a unified tool architecture, agents can execute reads, writes, and rollbacks via single-command sequences.
By leveraging JSON-based array formatting, this integration ensures that headings, code blocks, checklists, toggles, and tables compile correctly without parsing errors. The capability allows you to construct structured documentation, team wikis, and multi-layered task trackers directly within your production Craft workspace.
craft_read with depth parameters or custom search terms.craft_write using JSON-formatted block arrays, avoiding formatting errors common to basic markdown strings.To deploy this workflow, ensure your environment is configured for MCP v2 execution. This integration operates directly with your production account, so valid API access tokens and space identifiers are required. You can test your connection with the following command:
craft_read: "connection info"
Ensure your AI coding agent has this configuration defined in its execution context to expose the craft_write, craft_read, and blocks_revert tools.
The integration organizes data by identifying every document through its underlying rootBlockId. Blocks are sequenced elements that support diverse styling properties. Below is the main element-to-input mapping schema:
| Element Type | Input Format | Styling Property |
|---|---|---|
| Headings | JSON textStyle: h1 to h4 |
Standard margins, double newline rules |
| Tasks | JSON listStyle: task |
Supports todo, done, or canceled states |
| Code Blocks | JSON type: code |
Requires language parameter (no triple backticks) |
| Toggles | JSON listStyle: toggle |
Indentation levels from 0 to 5 define nesting |
| Collections | JSON database schema | Map custom properties using snake_case top-level keys |
All elements accept custom block styling such as hex background colors, alignment options, and decoration settings (e.g., blockquotes and callouts.
blocks_revert alongside target stamp validation to roll back accidental mutations safely within Openclaw Skills integrations.Loading
PuzzleTide Crossword helps Openclaw Skills generate valid, numbered crossword grids locally from custom clues, themes, or word banks.

Openclaw Skills for generating, running, and grading reproducible puzzle benchmarks that verify LLM and agent reasoning locally.

Openclaw Skills' Stock Analysis Expert turns A-share market data into a nine-layer investment report with scores, signals, and risk controls.

Openclaw Skills for Grok Build turns xAI's terminal AI coding agent into a full-screen TUI for interactive coding, file edits, shell automation, search, and headless workflows.

Openclaw Skills turns TRPG YAML map definitions into clean, readable SVG visual maps without modifying the source YAML.

Openclaw Skills for migrating and standardizing briefing_package map systems by moving large YAML maps to maps/, wiring map_ref, and enabling map_config visuals.








































