An intelligent utility to extract clean project hierarchies, full codebases, and snippet fragments from shared DeepSeek chat links using headless browser rendering.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deepseek-dev-assistant
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 deepseek-dev-assistant 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 DeepSeek Dev Assistant?
DeepSeek Dev Assistant is a robust development agent designed to bridge the gap between conversational AI ideation and active local software environments. Built as a specialized capability within the Openclaw Skills ecosystem, this tool automates the tedious task of translating chat histories into runnable source files. When triggered by a shared DeepSeek link or conversation query, it spins up a headless browser instance to handle the JavaScript-heavy layout of the share page, parses and strips conversational fluff, and organizes code blocks into their proper project tree location.
Rather than forcing you to copy-paste multiple iterations manually, this skill works recursively backward from the final message turn to reconstruct your workspace with the most up-to-date edits. By standardizing file writing and generating transparent task lists for partial fragments, this assistant accelerates continuous development pipelines using previous chat contexts as the authoritative source of truth.
DeepSeek Dev Assistant Use Cases
- Resuming complex programming tasks initialized in web-based DeepSeek chat threads directly inside local IDE environments.
- Syncing multi-turn iterative code changes and structural updates from online AI sessions with zero manual file creation.
- Scaffolding complete directory layouts and tracking down missing workspace files using the chat's generated ASCII project tree.
- Collating and resolving code fragments, partial functions, and length-constrained snippets inside an organized roadmap document.
How DeepSeek Dev Assistant Works
- Trigger Recognition: The assistant activates upon identifying a shared DeepSeek URL (
chat.deepseek.com/share/*) or matching input phrases. - Headless Page Loading: It initializes a browser-level fetch task using Puppeteer to fully render and load the dynamically populated React elements of the shared conversation.
- Boilerplate Trimming: Raw outer text is processed to strip platform chrome headers, copy buttons, and standard system disclaimers.
- Turn Priority Analysis: Working backwards from the last message to the first, it detects updated files and schedules only the most recent iteration of each file for write actions.
- Tree Diagnostics: An ASCII tree search isolates the target project outline to serve as a manifest of expected versus delivered source files.
- Deduplication & Categorization: Complete files are prepared for their target paths, while incomplete code fragments are compiled separately.
- Local Synthesis: Files are safely written to the active repository, accompanied by a dynamic todo list and a comprehensive project status overview file.
DeepSeek Dev Assistant Setup
To leverage this skill, ensure Node.js is configured on your environment and Puppeteer is active under the corresponding workspace directory.
1. Install Puppeteer Browser Dependency
Navigate to your local browser skill folder and confirm Puppeteer is installed before executing fetches:
cd ~/.openclaw/workspace/skills/browser && node -e "require('puppeteer')" 2>&1 || npm install puppeteer
2. Manual Testing Execution
Verify that the browser rendering layer loads and extracts content correctly from target share links:
cd ~/.openclaw/workspace/skills/browser && node index.js read "<your-deepseek-share-url>" 2>/dev/null
Once configured, invoking the agent via open-ended prompts containing share URLs will trigger the pipeline automatically.
DeepSeek Dev Assistant Data Schema & Taxonomy
During extraction, the skill builds dynamic configuration states and exports structural files as detailed below:
| Output File Name | Format | Description / Intended Target |
|---|---|---|
README.md |
Markdown | Summarizes the extraction source, timestamp, overall stack, completed assets, and missing file lists. |
FRAGMENTS_TODO.md |
Markdown | Collects partial snippets marked with labels like "关键代码片段" to avoid corrupting core source structures. |
memory/YYYY-MM-DD.md |
Markdown | Chronologically archives extraction parameters and actions to persist context across continuous Openclaw Skills agent runs. |
| Target Codebases | Directory Structure | Generated matching source files (such as Swift, Rust, HTML, etc.) based on structural ASCII tree rules. |
DeepSeek Dev Assistant Advanced Features
- Dynamic Reverse Parsing: Automatically isolates and prefers code files created in the latest turns of conversational chat to prevent regressions from earlier code iterations.
- Heuristic Page-Render Verification: Includes custom timing and size-threshold audits that automatically re-render JS-heavy share pages if early loads return incomplete outputs.
- Boilerplate Cleanup Filters: Employs deep text cleaning logic targeting platform markers, preventing IDE clutter and syntax errors from copy-button indicators or UI text.
- Tree Manifest Validation: Cross-checks extracted physical files against detected directory trees in the text to alert developers to code blocks missed due to LLM context output constraints.
SKILL.md
Loading
Related Openclaw Skills
Content Publisher Pro
An automated multi-platform content publishing tool designed to compile, differentiate, and deploy Markdown articles to GitHub Pages and Dev.to with built-in SEO optimization.
PopaDex Finance Copilot
A read-only personal finance copilot that provides comprehensive net worth insights, retirement planning, tax comparisons, and portfolio performance analysis.
Buttondown Integration
A secure integration enabling automated management and safety-first previewing of Buttondown newsletter drafts.
Codex Adapter Toolkit
A lightweight multi-provider API adapter for Codex that enables seamless switching, failover, and local hosting across multiple LLM providers.
NotFair Meta Ads Agent
An AI-powered assistant designed to safely diagnose, audit, and manage Facebook and Instagram advertising campaigns through NotFair's hosted MCP server.
Talos Lite
A free, automated content preview tool that generates a 1-week, 3-post social media schedule customized for a single platform based on your brand topic.
Featured*
