Deep Writer is an adaptive workflow skill designed to transform messy notes, transcripts, or topics into high-quality, structured long-form content through a staged editorial process.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install cell-deep-writer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install cell-deep-writer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Deep Writer?
Deep Writer is a sophisticated writing engine for Openclaw Skills that prioritizes depth, logic, and structural integrity over quick, superficial rewrites. It specializes in Chinese-language content creation, offering three distinct modes: source-driven for messy materials, draft-deepening for refining existing work, and topic-driven for building from a thesis. Unlike standard AI writers, it operates through a series of human-in-the-loop checkpoints to ensure the final output aligns perfectly with the target audience and purpose.
By utilizing specialized templates for briefs and outlines, Deep Writer ensures that every piece of content is grounded in a clear logic chain. It effectively filters out spoken-language noise, merges redundant ideas, and fills evidence gaps, making it an essential tool for researchers, journalists, and professional writers who need to produce publishable material from fragmented inputs.
Deep Writer Use Cases
- Converting raw interview transcripts or meeting notes into polished, professional articles.
- Deepening a rough draft by identifying weak logic, thin arguments, or missing supporting evidence.
- Generating structured long-form essays or reports from a single thesis, question, or outline.
- Cleaning up spoken-language fillers and repetitive ideas while preserving the original author's stance.
- Creating reusable content blueprints and detailed writing briefs for complex projects.
How Deep Writer Works
- Mode Identification: The skill analyzes the input to select between source-driven, draft-deepening, or topic-driven workflows.
- Analysis and Planning (Stage 1): It builds a comprehensive brief, identifies the audience/purpose, and clusters argument units into a logic chain.
- Structure Design (Stage 2): An optimized framework is designed based on the logic chain, merging duplicates and assigning specific 'jobs' to each section.
- Checkpoint Confirmation: The agent pauses for user approval of the brief and the structure before proceeding to full-scale drafting.
- Content Expansion (Stage 3): The full article is written, focusing on deepening claims and removing linguistic residue while adhering strictly to the locked structure.
- Quality Review: A final run against a quality-bar checklist ensures the piece is publishable and free of hallucinations.
Deep Writer Setup
To use Deep Writer within your environment, ensure the skill files are placed in your skills directory. You can initiate the workflow by providing your source material or topic to the agent.
# Example: Initializing a source-driven writing task
# Provide your transcript or notes and specify the goal
openclaw run deep-writer --input ./my-notes.md
Deep Writer Data Schema & Taxonomy
Deep Writer utilizes a structured asset and reference system to maintain consistency across the writing lifecycle:
| Asset Type | File Path | Purpose |
|---|---|---|
| Brief Template | assets/brief-template.md |
Standardizes the target audience, stance, and purpose definition. |
| Outline Template | assets/outline-template.md |
Provides a blueprint for the structural design phase. |
| Output Wrapper | assets/stage-output-template.md |
Manages the staged progress indicators and 'To-Do' status updates. |
| Quality Bar | references/quality-bar.md |
Contains the final review checklist and cleanup rules. |
| Workflow Guide | references/workflow.md |
Defines the heuristics for clustering and argument design. |
Deep Writer Advanced Features
- Multi-Mode Flexibility: Automatically switches logic based on whether the input is a messy transcript or a formal thesis.
- Staged Checkpoints: Prevents 'one-shot' failures by requiring user confirmation at critical logic and structure junctions.
- Logic Graphing: Extracts and visualizes the underlying argument chain to ensure coherent transitions and no circular reasoning.
- Source Preservation: In source-driven mode, it specifically protects the author's macro-frame while cleaning up micro-level noise.
- Gap Detection: Proactively identifies evidence gaps or assumptions that might weaken the final article.
SKILL.md
Loading
Related Openclaw Skills
ShareOne Skill
A streamlined skill for AI agents to host and share generated documents via persistent public URLs with optional security features.
V.A.R. Protocol
A professional safety and versioning protocol for AI agents to manage non-destructive updates and reliable asset recovery.
Hotspot Radar
A cross-platform hot search aggregator and trend analysis tool for real-time social media monitoring.
Burmese Audio Understanding
A high-accuracy transcription tool for Burmese audio powered by the Gemini 3.1 Flash Preview model.
Apple Ads CLI
A powerful read-only command-line interface for querying Apple Search Ads Campaign Management API data and performance reports.
Weekly Report Generator
An AI-driven automation tool for Openclaw Skills that aggregates GitHub activity, calendar events, and task lists into polished weekly reports.
Featured*
