Summarize for Openclaw

A high-speed CLI tool designed to extract and summarize content from URLs, YouTube transcripts, and local documents using various LLM providers.

utromaya-code
v1.0.0
Mar 23, 2026
0
2.5k
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install content-summarizer

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 content-summarizer 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 powerful command-line interface that bridges the gap between raw web content and actionable intelligence. As a key component of the Openclaw Skills ecosystem, it allows developers and AI agents to ingest information from diverse sources—including complex JavaScript-heavy websites, YouTube videos, and local PDF or text files—and generate concise, meaningful summaries.

By leveraging advanced extraction techniques like Firecrawl for blocked sites and Apify for YouTube fallbacks, this skill ensures that your AI agent always has access to the data it needs. It supports a wide array of models from providers like OpenAI, Anthropic, xAI, and Google, making it an adaptable solution for any research or automation workflow.

Summarize Use Cases

  • Distilling long-form articles or technical documentation into digestible bullet points.
  • Extracting best-effort transcripts from YouTube videos without needing heavy dependencies like yt-dlp.
  • Processing local PDF or text files to retrieve key findings during a coding or research session.
  • Automating content intelligence for AI agents using Openclaw Skills to understand external links provided by users.

How Summarize Works

  1. The user provides a source, which can be a URL, a YouTube link, or a local file path.
  2. The skill determines the optimal extraction method, utilizing internal parsers or external services like Firecrawl or Apify if configured.
  3. Extracted text is processed based on requested flags, such as length constraints or extraction-only modes.
  4. The content is sent to the configured LLM (defaulting to Gemini 3 Flash) to generate the final summary or JSON output.
  5. The results are returned to the terminal or the calling AI agent for further action.

Summarize Setup

To add this to your Openclaw Skills, install the binary via Homebrew:

brew tap steipete/tap
brew install summarize

Configure your environment with the necessary API keys:

export GEMINI_API_KEY="your_key_here"
# Optional: export FIRECRAWL_API_KEY="your_key"
# Optional: export APIFY_API_TOKEN="your_token"

Summarize Data Schema & Taxonomy

The skill can output data in standard text or structured JSON format. Configuration is managed via a central JSON file:

Feature Location / Format
Config File ~/.summarize/config.json
JSON Output Use the --json flag for machine-readable schemas
Default Model google/gemini-3-flash-preview
Supported Lengths short, medium, long, xl, xxl, or specific character counts

Summarize Advanced Features

  • Multi-provider support including OpenAI, Anthropic, xAI, and Google Gemini.
  • Intelligent fallback mechanisms using Firecrawl for bypassing bot detection on websites.
  • YouTube transcript extraction with Apify integration for high-reliability video data retrieval.
  • Granular output control with token limits and length presets to fit into AI context windows.
  • Headless automation compatibility for integration into larger Openclaw Skills workflows.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*