ghostfetch for Openclaw

A lightweight CLI tool for AI agents to search the web, fetch content as clean Markdown, and extract links without a full browser.

neothelobster
v1.1.0
Feb 27, 2026
1
1k
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install ghostfetch

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 ghostfetch 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 ghostfetch?

ghostfetch is a high-performance command-line utility designed to give LLM agents a streamlined way to interact with the web. As a key component in the Openclaw Skills ecosystem, it allows agents to perform web searches across major engines like DuckDuckGo, Google, Brave, and Bing without the overhead of a headless browser. It utilizes browser-like TLS fingerprints to ensure reliable access to websites while remaining a single, portable binary.

The primary value of ghostfetch lies in its ability to convert raw web content into clean, reader-mode Markdown. This process significantly reduces token consumption for AI models by stripping away unnecessary HTML boilerplate, styles, and scripts. By integrating this tool into your workflows, you provide your agents with a fast, secure, and token-efficient way to browse and synthesize information from the live web.

ghostfetch Use Cases

  • Performing multi-engine web searches to gather diverse perspectives for research tasks.
  • Fetching web pages as cleaned Markdown to provide context to LLMs while saving tokens.
  • Extracting specific sets of links from a page using regex filtering for automated crawling.
  • Gathering structured JSON data from web endpoints for data analysis or scraping.
  • Running parallel fetches on multiple URLs to speed up information gathering processes.

How ghostfetch Works

  1. The AI agent executes a ghostfetch command specifying either a search query or a direct URL.
  2. The tool selects the desired search engine or establishes a connection to the target site using a specific browser fingerprint (Chrome or Firefox).
  3. For search queries, ghostfetch parses the search engine results page and returns a list of URLs and snippets.
  4. For page fetches, the tool retrieves the HTML and, if the markdown flag is used, processes it through a reader-mode algorithm to extract main content.
  5. The tool outputs the result to stdout in the requested format (Text, Markdown, or JSON) for the agent to process further.

ghostfetch Setup

To add ghostfetch to your Openclaw Skills toolset, you must build the binary from the source and place it in your path. Ensure you have Go 1.21+ installed.

git clone https://github.com/neothelobster/ghostfetch.git
cd ghostfetch
go build -o ghostfetch .
cp ghostfetch ~/.openclaw/workspace/tools/

Alternatively, you can use the provided setup script within the repository to automate the installation and verification process.

ghostfetch Data Schema & Taxonomy

ghostfetch generates structured output that agents can easily parse. The output format is determined by command-line flags.

Output Type Format Content
Search List/JSON URL, Title, and Snippet for each result.
Fetch Markdown Cleaned reader-mode text or full page markdown.
Links List URLs extracted from the target page, optionally filtered by regex.
Structured JSON Comprehensive data including status codes, headers, and body content.

ghostfetch Advanced Features

  • Support for multiple search providers including DuckDuckGo, Brave, Bing, and Google.
  • Parallel fetch capabilities to handle multiple URLs simultaneously with configurable limits.
  • TLS fingerprinting to emulate specific browsers and avoid basic automated access blocks.
  • Captcha service integration for 2captcha and anticaptcha to bypass verification hurdles.
  • Cookie management options to maintain or disable persistence between requests.
  • Reader-mode extraction to isolate primary article content from cluttered web pages.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*