菜单
Analyze web and backend performance using Web Vitals, network timing, and Node.js metrics. Use when the user asks about page performance, load times, Core Web Vitals (LCP, CLS, INP), slow pages, backend bottlenecks, or SEO performance factors.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
CLI for driving an Android emulator (AVD) over adb. Use when the user needs to connect to an emulator (attach to a running one or boot/manage an AVD), tap/swipe/type on the UI, read the UI/accessibility tree, take screenshots or screen recordings, capture logcat, capture HTTP(S) requests in-process (Frida OkHttp hook — no proxy/CA, pinning bypassed), mock/intercept HTTP responses, pin a W3C trace id across app API calls, measure app performance, wait for network idle, or compare a screen against a Figma design. Emulator-only (no physical devices, no on-device tracepoints); network capture/mocking needs a rootable emulator image. Requires daemon; connect a device before other commands. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
CLI for backend service verification — drive HTTP/1.1+HTTP/2, gRPC, GraphQL, and WebSocket endpoints; capture logs from file, Docker, and Kubernetes sources; and verify state against Postgres, MySQL, or SQLite databases with snapshot/diff and change-feed primitives. Use when the user needs to call backend APIs over any of those protocols, assert on log lines or database rows, replay captured curl/HAR requests, manage session-scoped cookies/default headers, or correlate flows with W3C trace ids. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
Automates browser interactions using Playwright for web testing, debugging, and automation. Use when the user needs to navigate websites, take screenshots, fill forms, click elements, extract page content (HTML/text), audit accessibility (ARIA/AX tree), measure Web Vitals performance, monitor console logs and HTTP requests, mock API responses, execute JavaScript in browser, inspect React components, compare UI with Figma designs, manage reusable scenarios, or perform non-blocking debugging with tracepoints, logpoints, and exception monitoring. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
CLI for debugging Node.js backend processes with non-blocking inspection. Use when the user needs to connect to Node.js processes (by PID, name, Docker, or port), set tracepoints/logpoints/exceptionpoints, capture call stacks and local variables, inspect console logs, capture the process's outbound HTTP requests (egress) with W3C trace-context injection, or run reusable scenarios. Requires daemon; connect before other debug commands. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
CLI for debugging running Python processes with non-blocking inspection via debugpy (Debug Adapter Protocol). Use when the user needs to attach to a Python process (already-listening debugpy, PID injection, process name, or Docker), set tracepoints/logpoints/exceptionpoints, capture call stacks and local variables, dump all thread stacks (deadlock/GIL diagnosis), read stdout/stderr logs, capture the process's outbound HTTP requests (egress) with W3C trace-context injection, or run reusable scenarios. Requires daemon and `debugpy` installed in the target's Python environment; connect before other debug commands. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
CLI for driving interactive terminal programs (shells, REPLs, full-screen TUIs) over a pseudo-terminal (PTY). Use when the user needs to spawn a CLI/REPL/TUI attached to a TTY, type commands and keystrokes (tmux key syntax), wait for expected output before sending more input, capture raw scrollback or the rendered screen grid, send process signals, resize the terminal, or run a one-shot command and read its output + exit code. The platform spawns the process itself (it does not attach to a pre-existing external process). For server APIs use the backend CLI; for Node.js process debugging use the node CLI. Also covers the creds-gated Jira and Linear issue-tracker domains: fetch/search issues, download attachments, attach evidence, comment, transition/set state, and report verification results back to the tracker.
| name | performance-audit |
| description | Analyze web and backend performance using Web Vitals, network timing, and Node.js metrics. Use when the user asks about page performance, load times, Core Web Vitals (LCP, CLS, INP), slow pages, backend bottlenecks, or SEO performance factors. |
| allowed-tools | Bash(ironbee-browser-devtools-cli:*), Bash(ironbee-node-devtools-cli:*) |
Analyze web page performance using Web Vitals and network timing metrics.
This skill activates when:
ironbee-browser-devtools-cli o11y get-web-vitals
ironbee-browser-devtools-cli --json o11y get-web-vitals
ironbee-browser-devtools-cli --json o11y get-web-vitals --wait-ms 3000
ironbee-browser-devtools-cli --json o11y get-web-vitals --include-debug
ironbee-browser-devtools-cli --json o11y get-http-requests
ironbee-browser-devtools-cli --json o11y get-http-requests --resource-type script
ironbee-browser-devtools-cli --json o11y get-http-requests --resource-type stylesheet
ironbee-browser-devtools-cli --json o11y get-http-requests --status '{"min":400}'
ironbee-browser-devtools-cli content take-screenshot --name "above-fold"
ironbee-browser-devtools-cli content take-screenshot --name "full-page" --full-page
| Metric | Good | Needs Work | Poor |
|---|---|---|---|
| LCP | ≤2.5s | 2.5-4s | >4s |
| INP | ≤200ms | 200-500ms | >500ms |
| CLS | ≤0.1 | 0.1-0.25 | >0.25 |
| TTFB | ≤800ms | 800-1800ms | >1800ms |
| FCP | ≤1.8s | 1.8-3s | >3s |
SESSION="--session-id perf-audit"
# 1. Navigate to page
ironbee-browser-devtools-cli $SESSION navigation go-to --url "https://example.com"
# 2. Wait for page to settle
ironbee-browser-devtools-cli $SESSION sync wait-for-network-idle
# 3. Get Web Vitals
ironbee-browser-devtools-cli $SESSION --json o11y get-web-vitals --wait-ms 2000
# 4. Analyze network requests
ironbee-browser-devtools-cli $SESSION --json o11y get-http-requests
# 5. Check for console errors
ironbee-browser-devtools-cli $SESSION --json o11y get-console-messages --type warning
# 6. Take screenshot for reference
ironbee-browser-devtools-cli $SESSION content take-screenshot --name "performance-audit"
# 7. Cleanup
ironbee-browser-devtools-cli session delete perf-audit
# Filter by type (one at a time: document, stylesheet, image, script, xhr, fetch, etc.)
ironbee-browser-devtools-cli --json o11y get-http-requests --resource-type script
ironbee-browser-devtools-cli --json o11y get-http-requests --resource-type stylesheet
ironbee-browser-devtools-cli --json o11y get-http-requests --resource-type image
ironbee-browser-devtools-cli --json o11y get-http-requests
# Look for requests with high timing values
When TTFB or API latency suggests backend bottlenecks, inspect the Node.js process:
# Connect to API/server process
ironbee-node-devtools-cli --session-id perf debug connect --pid 12345
# Tracepoint on slow handler to capture call context
ironbee-node-devtools-cli --session-id perf debug put-tracepoint \
--url-pattern "routes/heavy.ts" \
--line-number 30