Summarize for Openclaw

A high-performance summarization engine that transforms URLs, documents, and multimedia files into actionable insights using Claude Opus 4.6.

evolinkai
v1.3.1
Mar 21, 2026
0
873
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install multi-summarize

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install multi-summarize 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 Summarize?

Summarize is a robust tool within the Openclaw Skills ecosystem designed for rapid content synthesis. Powered by evolink.ai and leveraging the state-of-the-art Claude Opus 4.6 model, it allows developers and researchers to extract key points from virtually any medium. Whether you are dealing with massive research papers or lengthy video tutorials, this skill distills information into structured takeaways.

The skill excels at cross-platform intelligence, handling everything from web scraping and PDF parsing to audio transcription. By integrating this into your Openclaw Skills workflow, you gain an automated research assistant capable of generating TL;DRs, meeting notes, and technical digests with human-level nuance and technical accuracy.

Summarize Use Cases

  • Rapidly digesting long-form technical documentation or blog posts from a URL.
  • Generating structured meeting notes or recaps from audio recordings and video files.
  • Extracting key takeaways and action items from dense PDF research papers or whitepapers.
  • Creating quick TL;DR summaries for large code files or markdown documentation.

How Summarize Works

  1. The user provides a source input such as a URL, local file path, or raw text to the Openclaw Skills interface.
  2. The system identifies the file type using MIME validation and resolves secure paths using realpath.
  3. Content is extracted using specialized binaries like pdftotext for documents or yt-dlp/whisper for media files.
  4. The extracted text is sent securely to the evolink.ai API for processing by the Claude Opus 4.6 model.
  5. The skill returns a structured Markdown response featuring a TL;DR, Key Takeaways, and Action Items.

Summarize Setup

First, ensure you have the required binaries installed on your system: curl, python3, realpath, and file. For media support, install yt-dlp and ffmpeg.

Configure your environment by setting your API key:

export EVOLINK_API_KEY="your-key-here"

You can then run the summarization script directly:

bash scripts/summarize.sh "https://example.com/article"

Summarize Data Schema & Taxonomy

The skill utilizes a tiered processing structure to manage data safely and efficiently:

Data Type Limit Processing Tool
Plain Text / MD 5MB Native Python/Bash
PDF Documents 50MB pdftotext / markitdown
Multimedia (Video/Audio) 1GB yt-dlp / whisper / ffmpeg

All temporary transcripts are stored in /tmp using PID-based filenames and are purged immediately after the summary is generated. Access is restricted to the SUMMARIZE_SAFE_DIR to ensure data sovereignty.

Summarize Advanced Features

  • Multi-model support: Override the default Claude Opus 4.6 by setting the EVOLINK_MODEL environment variable.
  • Security Hardening: Built-in blacklist for sensitive files like .env, SSH keys, and system configuration files.
  • Automated Media Transcription: Seamlessly converts video and audio into text before summarization.
  • Workspace Isolation: Strict path resolution prevents directory traversal and prefix-bypass attacks within Openclaw Skills.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*