What does this skill do, and when should you use it?
This skill connects to VS Code Insiders through the Chrome DevTools Protocol and uses @playwright/cli to operate the Copilot Chat extension and VS Code UI. It covers extension compilation, launch configuration, CDP attachment, snapshots, element interaction, and screenshot evidence. It also documents Monaco editor input handling, multiple webview targets, and restarting after extension code changes. It fits developers and QA engineers testing real VS Code interfaces, provided they use VS Code Insiders, compile the extension first, and maintain an authenticated GitHub session.
Runs npm install, npm run compile, or npm run watch; launches VS Code with code-insiders in extension-development mode, a CDP port, and a persistent user-data directory; attaches with npx @playwright/cli; lists or selects tabs, takes snapshots, fills inputs, sends key presses, and evaluates Monaco content; saves timestamped screenshots; and recompiles, terminates, and relaunches VS Code after source changes.
- An extension developer needs to load and verify an extension UI in VS Code Insiders.
- A Copilot Chat tester needs to fill the chat input and wait for generation to finish.
- A developer needs to discover VS Code UI elements through accessibility snapshots and interact with them.
- A QA engineer needs screenshots before and after UI actions as a visual test record.
- A tester must enter text into a Monaco editor when ordinary click-based input is blocked.
How do you install this skill?
- The persistent debug directory can retain Copilot/GitHub authentication state; verify its permissions and avoid screenshots containing tokens, source code, prompts, or personal data.
- The CDP port and internal CSS selectors introduce local exposure and version fragility; restrict the listener, verify the target window, and clean up processes and sensitive state afterward.
- Copilot authentication and model services may be affected by mainland-China network access, account policy, or regional restrictions; no offline or substitute path is documented.
- The cleanup example uses command-line matching to kill a process and could match the wrong VS Code instance; manually verify the process before terminating it.
- Shell / CLI
- Local filesystem
npm@playwright/cliVS Code Insidersauthenticated GitHub session
From the root of the microsoft/vscode repository, run npm install so @playwright/cli is available through devDependencies; the source also allows npm install -g @playwright/cli. The skill is located at extensions/copilot/.agents/skills/launch/SKILL.md. The supplied material does not document a separate installation or copy step.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/microsoft/vscode.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/extensions/copilot/.agents/skills/launch" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- automate VS Code,
- interact with chat,
- test the UI,
- take a screenshot,
Run npm run compile, then launch with code-insiders --extensionDevelopmentPath="$PWD" --remote-debugging-port=9223 --user-data-dir="$PWD/.vscode-ext-debug". Attach with npx @playwright/cli attach --cdp=http://127.0.0.1:9223, then use npx @playwright/cli tab-list and npx @playwright/cli snapshot. Example trigger prompts include “automate VS Code,” “interact with chat,” “test the UI,” “take a screenshot,” and “launch with debugging.”
What are this skill's strengths and limitations?
- Provides an end-to-end compile, launch, CDP attachment, snapshot, and interaction workflow.
- Documents reliable Monaco input methods using fill, type, press, and JavaScript focus events.
- Includes tab management, screenshot evidence, troubleshooting, and restart procedures.
- Uses a persistent profile so authentication, settings, and extension state can survive sessions.
- Requires VS Code Insiders; VS Code Stable will not activate the extension.
- Requires a compiled extension and an authenticated GitHub session for Copilot Chat to function.
- Relies on internal VS Code CSS selectors that may change after updates.
- The source does not provide a standalone package, test suite, or cross-platform validation results.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| VS Code Insiders UI Automation this page ✓ Microsoft · Official | 46 · Use with care | ★ 194k | 11d ago | MIT |
| Code OSS Launch & Debug ✓ Microsoft · Official | 54 · Use with care | ★ 194k | 11d ago | MIT |
| e2e: Agentic End-to-End Testing | 52 · Use with care | ★ 10k | 11d ago | Apache-2.0 |
| VS Code Chat Customizations Editor Skill ✓ Microsoft · Official | 52 · Use with care | ★ 194k | 11d ago | MIT |
| Full Page Screenshot Skill | 52 · Use with care | ★ 24 | 16d ago | — |
Compared with VS Code Stable, this skill requires VS Code Insiders because the target extension uses 58 proposed VS Code APIs and targets a specified version range.
How did FollowSkills review this skill?
The document constrains the declared tool pattern to Playwright CLI Bash usage and explains local CDP, an isolated user-data directory, and screenshot evidence. However, it persists authentication state, controls a chat UI, and may capture sensitive content without explicit per-action confirmation, data-flow disclosure, permission boundaries, sensitive-data guidance, or dependable rollback, so points are deducted.
The workflow documents prerequisites, dependencies, ports, retries, re-snapshotting, and common failures. Static material does not establish executable success on key paths; internal selectors, authentication, platform differences, and failure diagnostics remain limited, so the score is kept below the static ceiling.
Triggers, target scenarios, and main capabilities are fairly clear for UI interaction, chat automation, extension testing, debugging, and screenshots. Non-fit boundaries are incomplete, Chinese-language support is not addressed, and core Copilot use depends on GitHub authentication and services whose mainland-China reachability may vary, so points are deducted.
The document is organized progressively across prerequisites, workflow, attachment, interaction details, troubleshooting, and cleanup, with concrete commands and dependency/version hints. It lacks a clear versioning policy, changelog, maintenance owner, and update path, while some installation and environment assumptions remain implicit, so points are deducted.
For a developer with VS Code Insiders, a working build environment, and Copilot authentication, the workflow plausibly covers launch, attachment, snapshots, interaction, screenshots, and cleanup. Static review cannot verify result completeness or success rate, and authentication, compilation, and manual recovery add cost, so the score remains limited.
The skill provides concrete commands, ports, selectors, and failure conditions, and the repository contains related chat-library and performance CI. Those workflows do not directly cover this skill's Playwright/CDP paths, and no independent reproduction or third-party execution evidence is supplied, so only limited credit is justified.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →