camofox-mcp_skill

This skill enables stealth browser automation with anti-detection browsing, session management, and resource extraction for OpenClaw agents.

Python

2.5k

GitHub Stars

5

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill camofox-mcp

_meta.json276 B
CHANGELOG.md398 B
manifest.yaml3.6 KB
setup.sh164 B
SKILL.md5.5 KB

Overview

This skill provides an anti-detection browser automation toolkit for OpenClaw agents over an HTTP MCP transport. It exposes 41 tools for navigation, interaction, extraction, downloads, profiles, session management, and stealth web search. The design focuses on snapshot-first observation, session persistence, and resilience against fingerprinting and CAPTCHAs.

How this skill works

The skill connects an OpenClaw agent to a stealth browser controller that runs as an HTTP MCP server. Agents create tabs, navigate pages, interact with elements, and capture accessibility snapshots or screenshots. Built-in workflows handle form fills, batch clicks, scoped resource extraction and downloads, while profile and cookie tools persist sessions and reduce re-authentication.

When to use it

Automating login, multi-step forms, or authenticated scraping where detection risk is high
Collecting links, images, PDFs and other resources from web pages with reduced block rates
Running browser-based web searches across multiple engines and summarizing results
Taking accessibility snapshots to feed agent reasoning and reduce token usage
Managing long-lived sessions and restoring profiles across runs

Best practices

Start the CamoFox browser and MCP HTTP transport before invoking tools
Prefer snapshot (accessibility tree) for reading page content to save tokens over full HTML
Use profile save/load and cookie import to persist authentication and avoid repeated logins
Sequence high-level batch workflows (navigate_and_snapshot, fill_form, batch_click) for reliable automation
Monitor server_status and get_stats frequently to detect connectivity or performance issues

Example use cases

Automate a login and scrape account-specific data, then save profile cookies for future runs
Run a stealth web_search across Google and Reddit, open top results in tabs, snapshot each and extract links
Crawl a product listing page, batch_download images/PDFs and return file metadata
Perform a multi-step form submission: fill_form, type_and_submit, wait_for_text, then snapshot results

FAQ

Yes. The skill requires the CamoFox browser and the MCP HTTP transport to be running before use.

How do I reduce token usage when reading pages?

Use the snapshot (accessibility tree) tool as the primary read method; it captures structured content with far fewer tokens than full HTML or OCR.

Can I persist authenticated sessions across runs?

Yes. Use save_profile and load_profile or import_cookies to persist and restore session cookies and profiles.