anthropics/skills180kwebapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
浏览器自动化
Drive native macOS apps via interceptor macos *: AX trees, background click/type/keys/drag/scroll, occluded or minimized window capture, browser chrome, URL bars, OS dialogs, Apple Events, trusted OS input, monitor/replay, overlays, vision, speech, NLP, JSC host utilities, Electron app web-content control via interceptor macos cdp, and in-process app runtime control via interceptor macos runtime. Background-first by contract; only app activate and open --activate move focus. Use for native apps, named browser app routing, window control, cross-app work, and system UI. Not for content inside browser pages.
把这段话发给 Claude Code、Codex 或 Cursor。智能体会先检查安全性,你确认后才安装。
读取 https://funcoding.ai/skills/hacker-valley-media/interceptor/interceptor-macos/install.md ,按里面的步骤帮我安装这个 Skill。
Agent-operator skill for the macOS surface of Interceptor. Use the interceptor macos * CLI to drive native macOS applications: AX trees, OS-level trusted input, capture / vision / speech / NLP / Apple Events, monitor-and-replay, overlays. Use interceptor macos cdp / interceptor macos cdp app for Electron app web contents, and interceptor macos runtime for in-process native app runtime control. For content inside a browser tab load interceptor-browser instead.
The macOS bridge is a Swift daemon launched as a LaunchAgent / .app bundle. It uses Apple frameworks including Accessibility, ScreenCaptureKit, AVFoundation, Speech and Vision. Some existing capture and runtime paths also use private CGS symbols or injected runtime access. Check the surface reference and actual capability result; public framework support does not guarantee every private path works on every OS release.
This installed skill is self-contained. Source checkouts also have AGENTS.md, but packaged users may only have the skill directory below /Library/Application Support/Interceptor/skills.
Installed extensions may add capabilities under
interceptor macos <prefix> <cmd>. Each carries its own skill (interceptor-ext-<name>); this skill does not describe them. Runinterceptor extensions listto see what is installed.
interceptor status # 1. Confirm daemon + bridge are alive
interceptor macos trust # 2. Confirm TCC permissions granted
interceptor macos open "Finder" # 3. Tree + windows (background — does NOT raise Finder)
interceptor macos read # 4. AX tree + frontmost info
interceptor macos act e5 # 5. AX press of ref e5 — no focus change
interceptor macos act e3 "hello" # 6. AX value-set of ref e3 — no focus change
Treat eN refs as short-lived. AX state can change between calls; re-read before acting.
Only two commands move focus: interceptor macos app activate <app> and interceptor macos open <app> --activate. Everything else is background-first by contract — open (without --activate), all input verbs, all reads, capture, AX, menu, intent dispatch, scroll, drag, vision, overlays. If you call any other command and the user's frontmost app changes, that is a bug — file it.
The user is on this Mac too. Assume the person is working while you work. Never run either focus command to make your own work easier or to test something. Use one only when the user asks for the app to come forward, or when the action truly cannot be done in the background, and then say so first, keep it to one fast batch, and put back the frontmost app and the cursor.
What is truly foreground-only today: launching an app that is not running (it can come forward even from open without --activate, and an app updater can raise a password prompt), and a click or drag on a minimized window or hidden app. Everything else has a background route: refs, --app / --pid / --window, and interceptor click <ref> --trusted for a page in an unfocused browser window.
Full contract + verb inventory + worked examples + pitfalls: references/background-first.md.
Each workflow is a complete self-contained "you are doing X" procedure. Open the file when the task matches.
| Workflow | When to invoke |
|---|---|
workflows/capture-backgrounded-app.md | Screenshot an occluded / minimized / cross-Space window — without activating it |
workflows/drive-backgrounded-app.md | Click / type / keys / drag / scroll against a non-frontmost app via AX refs or window-addressed delivery (--app, --window) |
workflows/dispatch-apple-event.md | Apple Events to a named bundle id — open URL in Brave, read active tab, etc. |
workflows/read-ax-tree.md | tree --app of any app, with automatic Electron wake-up |
workflows/record-and-replay-mac-flow.md | macos monitor record + export + replay native UI flows |
workflows/trusted-input-gate.md | Satisfy an OS-level trusted-input gate that filters synthetic CGEvents |
workflows/clear-human-verification-gate.md | Clear a CAPTCHA / human-verification gate (reCAPTCHA, Turnstile, hCaptcha, generic) in the user's own signed-in session with a trusted click delivered to the browser window in the background |
| File | Topic |
|---|---|
references/background-first.md | Full Background-First contract, verb inventory, reflexes-to-drop, pitfalls |
references/accessibility-and-input.md | AX tree mechanics, input routing, window control, credential delivery by name (--secret) |
references/capture-and-vision.md | ScreenCaptureKit + CGS capture, Vision OCR, audio intelligence |
references/advanced-domains.md | Apple Events, container runtime, OS log, fs, URL fetch, file watch |
references/jsc-host.md | Plain JavaScriptCore host capabilities, flags, scope rules, and pitfalls |
references/monitor-and-replay.md | Native monitor sessions, replay plans, event sources |
references/command-catalog.md | Full macOS command surface with flags and examples |
references/permissions.md | TCC permissions, microphone re-poll, Dock-icon notes |
references/cdp-app.md | Electron/Chromium desktop apps via interceptor macos cdp / interceptor macos cdp app |
references/native-agent.md | In-process native app runtime control via interceptor macos runtime |
If the target is inside a browser page (DOM, network, SPA state, browser monitor, scene graph of a rich editor) - load interceptor-browser instead.
If the target is inside an Electron/Chromium desktop app's web contents (Slack, VS Code, Notion, Descript, etc.) - use references/cdp-app.md, not AX screenshots.
If the target is inside a native app's runtime (AppKit/SwiftUI object graph, selector calls, rendered text changes, hooks, MapKit state) - use references/native-agent.md. AX is still the fallback projection when the in-process agent is unavailable.
anthropics/skills180kToolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
浏览器自动化
addyosmani/agent-skills103kTests in real browsers via Chrome DevTools MCP. Use when building or debugging anything that runs in a browser. Use when you need to inspect the DOM, capture console errors, analyze network requests, profile performance, or verify visual output with real runtime data. Requires the chrome-devtools MCP server to be configured.
浏览器自动化
ComposioHQ/awesome-claude-skills77kToolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
浏览器自动化
code-yeongyu/oh-my-openagent70kDrives a real browser through the omowright library from the js eval kernel: sites the user is already signed into, forms and clicks, JS-rendered pages, screenshots, web QA, extension popups, a human handoff for login, CAPTCHA or OTP, and a browser you own for scraping, bot-scored targets, network capture and QA traces. Use for any interactive browser task; not for a plain search or an unblocked static fetch.
浏览器自动化
shanraisshan/claude-code-best-practice67kBrowser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
浏览器自动化
CherryHQ/cherry-studio52kRun Cherry Studio critical-path system regression tasks through the repository-owned Playwright E2E workflow. Use for full regression, release acceptance, development-branch system validation, or a named cherry-regression-test task on GitHub-hosted macOS and Windows runners.
浏览器自动化