Build AI chat interfaces using ai-elements components — conversations, messages, tool displays, prompt inputs, and more. Use when the user wants to build a chatbot, AI assistant UI, or any AI-powered chat interface.
中文简介正在准备中。当前显示原文说明。
Chrome DevTools CLI for browser automation via shell commands. Use when interacting with web pages from the command line — navigating, clicking, filling forms, inspecting console/network, taking screenshots, or extracting page content. Triggers on: browse a page, automate Chrome, inspect console, check network requests, take a screenshot, fill a form, click a button.
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
CLI for Chrome DevTools Protocol. All list commands output NDJSON (one JSON object per line) — grep/tail/jq friendly. Requires Chrome running with remote debugging.
Chrome must be running with --remote-debugging-port:
cdp-cli launch # macOS: launches Chrome with debugging on :9223
Or start Chrome manually with --remote-debugging-port=9222 and pass
--cdp-url http://localhost:9222.
All commands taking <page> accept a page ID or a title substring.
Use cdp-cli tabs to discover pages.
cdp-cli tabs # List all open pages (NDJSON)
cdp-cli tabs | grep "example" # Find by title
cdp-cli is designed to minimize token usage by default:
--with-type, --with-timestamp, --with-source only when
needed. Use -v for all fields.--format ax) which is compact.
Use --format text for plain text, --format dom for full DOM (heavy).--duration.Strategy: Start with defaults, add verbosity flags only when needed.
cdp-cli tabs # List all pages (NDJSON)
cdp-cli new "https://example.com" # Open new tab
cdp-cli new # Open empty tab
cdp-cli go "<page>" "https://url.com" # Navigate to URL
cdp-cli go "<page>" back # History back
cdp-cli go "<page>" forward # History forward
cdp-cli go "<page>" reload # Reload
cdp-cli close "<page>" # Close tab
cdp-cli snapshot "<page>" # Accessibility tree (default, compact)
cdp-cli snapshot "<page>" -f text # Plain text content
cdp-cli snapshot "<page>" -f dom # Full DOM tree (JSON, heavy)
cdp-cli eval "<page>" "document.title" # Evaluate JS expression
cdp-cli eval "<page>" "Array.from(document.querySelectorAll('h1')).map(h => h.textContent)"
Use the accessibility tree to discover elements by role/name, then construct CSS selectors for click/fill.
cdp-cli console "<page>" # Last 10 messages, bare strings
cdp-cli console "<page>" -n 50 # Last 50 messages
cdp-cli console "<page>" --all # All messages (or -n -1)
cdp-cli console "<page>" -t error # Filter by type (log/error/warn/info)
cdp-cli console "<page>" -d 2 # Collect for 2 seconds
cdp-cli console "<page>" -i # --inspect: expand nested objects/arrays
cdp-cli console "<page>" -v # --verbose: all fields (type, timestamp, source)
cdp-cli console "<page>" --with-type # Add type field only
cdp-cli console "<page>" --with-source # Add source location (url, line)
When truncated, stderr shows: (N messages skipped. Use --tail M or --all to see more)
cdp-cli network "<page>" # Requests in last 0.1s (NDJSON)
cdp-cli network "<page>" -d 5 # Collect for 5 seconds
cdp-cli network "<page>" -t fetch # Filter by type (xhr, fetch, script, etc)
cdp-cli network "<page>" -d 5 -t fetch # Combine filters
Filter results with grep:
cdp-cli network "<page>" | grep '"status":4' # 4xx errors
cdp-cli network "<page>" | grep '"status":404' # 404s specifically
cdp-cli click "<page>" "button" # Click by CSS selector
cdp-cli click "<page>" "button" -d # Double click
cdp-cli click "<page>" "button" -g # --user-gesture: required for
# WebXR, fullscreen, and other
# activation-gated APIs
cdp-cli click "<page>" --node 42 # Click by backend DOM node ID
cdp-cli click "<page>" "button" -w 5000 # Wait up to 5s for selector
cdp-cli fill "<page>" "user@example.com" "input[name='email']"
cdp-cli fill "<page>" "secret" "input[type='password']"
cdp-cli key "<page>" enter
cdp-cli key "<page>" tab
cdp-cli key "<page>" escape
cdp-cli screenshot "<page>" screenshot.jpg # Save JPEG (default)
cdp-cli screenshot "<page>" out.png -f png # PNG format
cdp-cli screenshot "<page>" out.webp -f webp # WebP format
cdp-cli screenshot "<page>" out.jpg -q 50 # Lower quality (0-100)
All errors output NDJSON with "error": true:
{ "error": true, "message": "Page not found: example", "code": "PAGE_NOT_FOUND" }
cdp-cli tabs | grep "example" # Find page
cdp-cli snapshot "example" # Get accessibility tree
# Identify selectors from the tree, then:
cdp-cli fill "example" "query" "input"
cdp-cli click "example" "button[type='submit']"
cdp-cli screenshot "example" result.jpg
cdp-cli console "localhost" -t error -d 10 # Errors over 10 seconds
cdp-cli console "localhost" -t error -v # With full source location
cdp-cli network "localhost" -d 5 | grep '"status":4' # Failed requests
cdp-cli network "localhost" -d 5 -t fetch # API calls only
cdp-cli new "http://localhost:3000/login"
cdp-cli fill "localhost" "user@example.com" "#email"
cdp-cli fill "localhost" "password123" "#password"
cdp-cli click "localhost" "button[type=submit]"
sleep 2
cdp-cli eval "localhost" "document.querySelector('.success')?.textContent"
| Option | Default | Description |
|---|---|---|
--cdp-url | http://localhost:9223 | Chrome DevTools Protocol URL |
--help | — | Show help |
--version | — | Show version |
还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Build AI chat interfaces using ai-elements components — conversations, messages, tool displays, prompt inputs, and more. Use when the user wants to build a chatbot, AI assistant UI, or any AI-powered chat interface.
中文简介正在准备中。当前显示原文说明。
Search the live web via Perplexity Search API. Use when you need current documentation, release notes, vendor pages, news, domain-constrained web search, or date/recency filtering. Not for local codebase search or stable docs already in context.
中文简介正在准备中。当前显示原文说明。
Uses the chrome-devtools-axi CLI for browser automation, accessibility-tree snapshots, console and network inspection, screenshots, Lighthouse audits, and performance traces. Use when interacting with Chrome from the shell, especially when the user mentions chrome-devtools-axi, AX snapshots, browser debugging, or DevTools automation from the command line.
中文简介正在准备中。当前显示原文说明。
Deep expertise in cmux — the terminal multiplexer with native browser views. Use when managing panes, reading terminal output, sending keystrokes, opening browser views, or manually testing web UIs and TUIs inside cmux. Triggers on: cmux, open a browser pane, split terminal, read screen, send keys, test this UI in cmux, preview in cmux.
中文简介正在准备中。当前显示原文说明。
Uses the gh-axi CLI for GitHub shell operations: issue, pull request, workflow run, release, repo, search, and API tasks. Prefer this over regular `gh` for GitHub reads and simple mutations when an agent needs compact, structured, suggestion-rich output. Triggers on: gh, GitHub CLI, github issue, github pr, pull request, workflow run, github release, gh api, repo inspection, list PRs, view issue, check workflow runs, inspect repo, GitHub shell operations.
中文简介正在准备中。当前显示原文说明。
Work with Graphite (gt) for stacked PRs. Use when creating, navigating, submitting, or managing PR stacks, or when the user mentions gt, Graphite, or stacked diffs.
中文简介正在准备中。当前显示原文说明。