What does this skill do, and when should you use it?
This skill handles installing a Visual Studio Code extension from an extension ID. It recognizes IDs that typically use the publisher.extensionName format and invokes the VS Code workbench.extensions.installExtension command. It installs the stable release by default, switching to a pre-release when the user explicitly requests one or the environment is VS Code Insiders. The skill is one of 61 skills bundled in microsoft/vscode, whose source code is licensed under MIT.
It accepts an extension ID such as ms-python.python, builds arguments in the form [extensionId, { enable: true, installPreReleaseVersion: boolean }], and runs workbench.extensions.installExtension through copilot_runVscodeCommand with skipCheck set to true. It does not describe an extension's features or install multiple extensions.
- A developer wants to add a capability to an existing VS Code setup and knows the extension ID.
- A user explicitly asks for a pre-release extension version.
- An extension must be installed in a VS Code Insiders environment.
- A user wants an agent to execute the VS Code installation command instead of installing the extension manually through the Extensions view.
How do you install this skill?
- Installing an extension changes the user's VS Code environment; the extension ID, version choice, and intent should be confirmed first.
- skipCheck=true bypasses command-existence checking; invalid IDs, unreachable networks, installation failures, and rollback are not specified.
- Extension provenance, permissions, supply-chain risk, and mainland-China network reachability are not documented by the skill.
Visual Studio Codecopilot_runVscodeCommand
The skill file is located in https://github.com/microsoft/vscode at extensions/copilot/assets/prompts/skills/install-vscode-extension/. The source does not document a standalone installation command, destination folder, or packaging procedure for a particular client; expose the file using the installation mechanism of the Agent Skills client in use.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/microsoft/vscode.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/extensions/copilot/assets/prompts/skills/install-vscode-extension" ~/.claude/skills/
rm -rf "$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:
- Install the ms-python.python extension in VS Code.
Use a request such as “Install the ms-python.python extension in VS Code.” The agent should call workbench.extensions.installExtension with arguments ["ms-python.python", {"enable": true, "installPreReleaseVersion": false}], execute it through copilot_runVscodeCommand, and set skipCheck to true. Set installPreReleaseVersion to true when the user explicitly requests a pre-release or the current environment is VS Code Insiders; otherwise set it to false.
What are this skill's strengths and limitations?
- Uses one clearly defined VS Code command and the standard extension ID format.
- Defines explicit stable-versus-pre-release behavior for ordinary VS Code and Insiders scenarios.
- Always requests the extension to be enabled, making the intended post-install state clear.
- Requires an environment that can execute copilot_runVscodeCommand; other environments may need tool-call adaptation.
- Covers installation by ID only, not extension search, removal, updating, or configuration.
- The source provides no test suite, failure-handling procedure, or cross-platform validation details.
- The source does not document a standalone installation procedure for the skill itself.
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 Extension Installer this page ✓ Microsoft · Official | 41 · Not recommended | ★ 194k | 11d ago | MIT |
| Callimachus Recall | 58 · Recommended | ★ 43 | 3mo ago | AGPL-3.0 |
| OpenChamber UI/API Decoupling Skill | 58 · Recommended | ★ 12k | 11d ago | MIT |
| Search View Results Reader ✓ Microsoft · Official | 43 · Not recommended | ★ 194k | 11d ago | MIT |
| agentMemory — Hybrid Memory System for AI Coding Agents | 42 · Not recommended | ★ 33 | 9mo ago | MIT |
How did FollowSkills review this skill?
The skill clearly installs and enables a specified extension, creating an external environment change; it lacks user confirmation, source or data-flow disclosure, rollback guidance, and explicitly requires skipCheck=true, so points are deducted.
The command, argument shape, and stable versus pre-release rule are clear for the happy path; tool availability, invalid extension IDs, network failures, and diagnostic feedback are unspecified. Static calibration limits the score to 10 or below.
The audience, scenario, and trigger are reasonably clear: install a VS Code extension by ID; non-fit cases, input validation, Chinese-language support, and mainland-China network reachability are not addressed, so points are deducted.
The document is concise and structured, while repository evidence supplies MIT licensing, versioning, and Microsoft maintenance provenance; the skill lacks dependency notes, output examples, limitations, troubleshooting, changelog, and an independent update path.
The core task has a direct VS Code command and tool invocation and plausibly completes installation; however, success output, confirmation, recovery on failure, and comparative alternatives are absent, limiting directly usable evidence. Static calibration limits the score to 7 or below.
The skill instructions and command path are auditable, and the repository contains CI and test infrastructure; the supplied CI/tests do not cover this installation path and no independent reproduction evidence is included, so points are deducted.
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 →