agentsclimarketplace
Sign in

Patchright cli

Skill AhaiMk01/patchright-cli/skills/patchright-cli

Anti-detect browser automation using Patchright (undetected Playwright fork). Use when you need to interact with websites that block regular Playwright/Chrome DevTools, such as Akamai/Cloudflare-protected sites. Provides the same command interface as playwright-cli but bypasses bot detection. Use this skill whenever the user asks to automate a browser, scrape a website, fill forms, log into sites, take screenshots, or do anything involving Chrome automation — especially if the target site has bot protection like Cloudflare or Akamai.From its SKILL.md

Install
npx -y skills add AhaiMk01/patchright-cli --skill patchright-cli

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

11.9 KB, as published. Nobody here has run it

Anti-Detect Browser Automation with patchright-cli

patchright-cli drives a real Chrome browser that passes bot detection (Cloudflare, Akamai, etc.). It works as a CLI: you issue commands, get back page state. This is the tool to reach for whenever regular Playwright or Chrome DevTools gets blocked.

Quick start

patchright-cli open https://example.com     # launch browser + navigate
patchright-cli snapshot                      # get element refs
# read the snapshot YAML to find the right ref, then:
patchright-cli click e5                      # interact by ref
patchright-cli fill e3 "search query"        # type into an input
patchright-cli press Enter                   # press a key
patchright-cli screenshot                    # capture the page
patchright-cli close                         # done

Every interaction follows: open -> snapshot -> read refs -> interact -> snapshot again -> repeat. Refs are ephemeral -- they change on every page update. If a ref fails, re-snapshot.

Command chaining

Commands can be chained with &&. The browser persists via a background daemon, so chaining is safe.

# Chain when you don't need intermediate output
patchright-cli open https://example.com && patchright-cli screenshot

# Run separately when you need to read refs first
patchright-cli snapshot          # read refs from output
patchright-cli click e5          # use discovered ref

Global options

These go before the command:

--headless              # Run headless (default: headed -- headed is less detectable)
--persistent            # Use persistent profile (keeps cookies/storage across sessions)
--profile=/path         # Custom profile directory
--proxy=<url>           # Proxy server (http, https, socks5) -- supports user:pass@host auth
-s=mysession            # Named session (default: "default", or PATCHRIGHT_CLI_SESSION env var)
--port=9322             # Custom daemon port (default: 9321)
--config=<path>         # Load options from JSON config file
--cdp=<url>             # Attach to Chrome via CDP endpoint (use with `attach` command)
--device="iPhone 15"    # Emulate a device
--viewport-size=1280x720 # Set viewport size
--locale=en-US          # Browser locale
--timezone=America/New_York # Timezone ID
--geolocation=40.7,-74.0   # Geolocation override (lat,lon)
--user-agent=<ua>       # Custom user agent string
--grant-permissions=geolocation,camera  # Grant permissions at launch
--timeout-action=10000  # Default action timeout (ms)
--timeout-navigation=30000 # Default navigation timeout (ms)
--show-port=9322        # Dashboard port (default: 9322)
--raw                   # Strip page/snapshot decorations from output
--json                  # Wrap full response as JSON ({success, output, ...})

Core commands

Browser lifecycle

open [url]                    # Launch browser (optionally navigate)
open --persistent             # Keep cookies/storage between runs
open --headless               # Headless mode
attach --cdp=<url>            # Attach to existing Chrome via CDP
close                         # Close session (closes the browser)
detach                        # Detach an attached session (keeps external browser running)

Navigation

goto <url>                    # Navigate to URL
go-back / go-forward          # History navigation
reload                        # Reload page
url                           # Print current URL
title                         # Print page title

Snapshots

snapshot                      # Full page snapshot
snapshot <ref>                # Subtree of a specific element
snapshot --depth=N            # Limit depth
snapshot -i                   # Interactive elements only
snapshot --boxes              # Append [box=x,y,w,h] to each [ref=eN] line

Interaction

click <ref>                   # Left-click (smart: navigates links directly)
click <ref> right             # Right-click
click <ref> --modifiers=Alt   # Click with modifier keys
dblclick <ref>                # Double-click
hover <ref>                   # Hover
drag <ref> <target-ref>       # Drag element to target
drop <ref> --path=./file.png  # Drop file onto element (HTML5 drop from outside)
drop <ref> --data="text/plain=hello"  # Drop arbitrary MIME data

Form input

fill <ref> "value"            # Clear field and type value
fill <ref> "value" --submit   # Fill and press Enter
type "text"                   # Type via keyboard (no target)
type "text" --submit          # Type and press Enter
select <ref> "option"         # Select dropdown option
check <ref> / uncheck <ref>   # Checkbox/radio

Text and screenshots

text <ref>                    # Get text content by ref
text "css-selector"           # Get text by CSS selector
screenshot                    # Page screenshot
screenshot <ref>              # Element screenshot
screenshot --full-page        # Full scrollable page
screenshot --filename=F       # Custom filename

Scroll, wait, and keyboard

scroll <dx> <dy>              # Scroll by pixels
scroll-to <ref>               # Scroll element into view
wait <ms>                     # Wait N milliseconds
wait-for <ref>                # Wait until element visible
wait-for <ref> --state=hidden # Wait until hidden
press <key>                   # Keypress (Enter, ArrowDown, etc.)
keydown <key> / keyup <key>   # Hold/release key

Mouse

mousemove <x> <y>             # Move to coordinates
mousedown / mouseup           # Button down/up
mousewheel <dx> <dy>          # Scroll wheel

Tabs

tab-list                      # List open tabs
tab-new <url>                 # Open new tab
tab-select <index>            # Switch to tab
tab-close [index]             # Close tab

Dialogs

Dialogs must be pre-armed before the action that triggers them:

dialog-accept                 # Accept next dialog
dialog-accept "OK"            # Accept with text input
dialog-dismiss                # Dismiss next dialog

DevTools

console                       # Show console messages
console warning               # Filter by level
network                       # Show network requests (each line is `#<id> METHOD STATUS URL`)
requests                      # Alias for `network`
network --static              # Include static assets
request <id>                  # Full detail for one request (headers + body)
request <id> --body           # Also fetch response body
generate-locator <ref>        # Emit Playwright locator string for a ref
highlight <ref>               # Persistent overlay on element [--style=CSS]
highlight <ref> --hide        # Remove overlay; `highlight --hide` clears all

Recording and capture

screenshot / pdf              # Static capture
video-start / video-stop      # Video recording
tracing-start / tracing-stop  # Playwright trace
codegen / codegen-stop        # Record interactions as script
resize <w> <h>                # Viewport resize
upload <file> [ref]           # File upload
show                          # Live dashboard

Storage

cookie-list / cookie-get / cookie-set / cookie-delete / cookie-clear
cookie-export / cookie-import
localstorage-list / localstorage-get / localstorage-set / localstorage-delete / localstorage-clear
sessionstorage-list / sessionstorage-get / sessionstorage-set / sessionstorage-delete / sessionstorage-clear
state-save <file> / state-load <file>

Network

route "<pattern>" [options]   # Mock requests (--status, --body, --content-type, --header)
route-list / unroute          # Manage routes
network-state-set offline     # Simulate offline
network-state-set online      # Restore connectivity

Sessions

list                          # List active sessions
close-all                     # Close all sessions
kill-all                      # Force-kill all + stop daemon
delete-data                   # Delete session profile data
grant-permissions <perms>     # Grant browser permissions

Common patterns

Login flow

patchright-cli open https://example.com/login --persistent
patchright-cli snapshot
patchright-cli fill e3 "[email protected]"
patchright-cli fill e5 "password123"
patchright-cli click e8
patchright-cli snapshot
# With --persistent, cookies survive across sessions

Extracting data

patchright-cli open https://example.com/data
patchright-cli snapshot
cat > /tmp/extract.js << 'JSEOF'
JSON.stringify(
  [...document.querySelectorAll('table tr')].map(row =>
    [...row.cells].map(cell => cell.textContent.trim())
  )
)
JSEOF
patchright-cli eval --file=/tmp/extract.js

Using a proxy

patchright-cli --proxy=http://host:port open https://example.com
patchright-cli --proxy=socks5://host:port open https://example.com
patchright-cli --proxy=http://user:pass@host:port open https://example.com

Handling dialogs

patchright-cli dialog-accept              # Arm before triggering action
patchright-cli click e5                   # This click triggers the dialog

Waiting for dynamic content

patchright-cli wait-for e12               # Wait for element
patchright-cli wait 1000                  # Fixed wait
# For custom conditions, use run-code (see references/running-code.md)

Anti-detect notes

  • Uses real Chrome (not Chromium) -- this is what makes it undetectable
  • Patchright patches navigator.webdriver and other detection vectors automatically
  • Headed by default -- headless mode is more detectable, use only when necessary
  • No custom user-agent or headers by default -- preserves Chrome's natural fingerprint
  • Persistent profiles maintain realistic browser history and cookies
  • The daemon architecture means Chrome stays running between commands, behaving like a real user's browser
  • attach --cdp lets you connect to an existing Chrome instance

Config files

Create .patchright-cli/config.json in your project directory (auto-discovered) or pass --config=<path>:

{
  "headless": false,
  "persistent": true,
  "proxy": "http://proxy:8080",
  "device": "iPhone 15",
  "locale": "en-US",
  "timezone": "America/New_York"
}

CLI flags always override config file values.

Installation

CLI tool

# Recommended -- always runs latest version
uvx patchright-cli <command>

# Or install globally
pip install patchright-cli

Skill (for AI coding agents)

# Install for all detected agents (Claude Code, Cursor, Gemini, Codex, 40+ more)
npx skills add AhaiMk01/patchright-cli

Other methods:

# Via patchright-cli itself
pip install patchright-cli && patchright-cli install --skills

# Or just tell your agent:
# "Install patchright-cli skill from https://raw.githubusercontent.com/AhaiMk01/patchright-cli/main/skills/patchright-cli/SKILL.md"

Specific tasks

Keep looking

Skills are one crate of 628,971. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.