Centrally manage VS Code settings with policies
Enterprise policies in Visual Studio Code enable organizations to centrally manage VS Code settings for their development teams to ensure consistency and compatibility across their organization. When a policy value is set, the value overrides the VS Code setting value configured at any level (default, user, and workspace).
IT admins can deploy and enforce specific VS Code configurations on users' devices through different device management solutions. VS Code supports applying policies on Windows, macOS, and Linux.
Go directly to your task:
- Deploy policies on Windows.
- Deploy policies on macOS.
- Deploy policies on Linux.
- Verify applied policies.
- Look up a policy.

In this article, you learn which enterprise policies are available in VS Code and how to configure them on different operating systems.
Choose policy values by following the guidance for extensions, AI features, telemetry, and updates. For AI controls, also review precedence and management scope.
Sample policy files are configuration examples, not a recommended organizational baseline. Before deployment, remove policies you don't intend to enforce and verify the remaining values on a test device.
Windows group policies
VS Code has support for Windows Registry-based Group Policy.
You can deploy policies through Active Directory Group Policy or a device management solution such as Microsoft Intune. Use the Local Group Policy Editor to test policies on an individual device.
Step 1: Obtain the sample ADMX and ADML files
Starting from VS Code version 1.69, each release ships with a policies directory containing ADMX template files that define the available policies.
You can get the ADMX and ADML files from either an existing installation or by downloading and extracting the VS Code zip archive. Follow these steps to obtain the files:
- Download the VS Code zip archive for your version of VS Code.
- Extract the zip file to a temporary location.
- Navigate to the
policiesfolder in the extracted files. This folder contains the ADMX template file (for example,VSCode.admx) and a folder for each language, such asen-us, with the ADML file for that language.
Step 2: Install or import the policy templates
Make the templates available to your management tool before configuring values:
- For Active Directory environments, copy the ADMX and language-specific ADML files to the Central Store.
- For cloud-managed devices, import the ADMX and ADML files into your management solution, then configure and assign the imported administrative template policies.
- For a local test, install the files as described below.
Install templates for a local test
- Copy the
VSCode.admxfile toC:\Windows\PolicyDefinitions. - Copy the ADML file from the folder for your language (for example,
en-us\VSCode.adml) toC:\Windows\PolicyDefinitions\<your-locale>(for example,C:\Windows\PolicyDefinitions\en-US).
You need administrator privileges to copy files to the PolicyDefinitions directory and a Windows edition that includes the Local Group Policy Editor.
Step 3: Configure policy values
For a domain deployment, edit the Group Policy Object that applies to your target devices or users. For a local test, run gpedit.msc to open the Local Group Policy Editor.
The policies are grouped by feature under:
- Computer Configuration > Administrative Templates > Visual Studio Code
- User Configuration > Administrative Templates > Visual Studio Code
Computer-level policies take precedence over user-level policies when both are configured.
For example, to turn off Visual Studio Code's auto-update:
- Under Visual Studio Code, open Update and then UpdateMode.
- Select Enabled, then choose
nonein the policy's value dropdown. - Select OK to save the changes.
For a Boolean policy, select Enabled and use the checkbox in the policy's options to set its value to true or false. The policy's Enabled state means that its configured value is enforced; it is not the value itself. Use Not Configured when you don't want the Group Policy Object to enforce that policy.
Configured values are written under Software\Policies\Microsoft\VSCode in the machine or user registry hive, depending on the scope you chose.
Step 4: Deploy and verify the policies
Apply the Group Policy Object or assign the policy configuration through your device management solution to a test group first. After the device receives the configuration, restart VS Code and follow Verify policy enforcement.
For the UpdateMode example, confirm that the report lists UpdateMode with the value none and that the application's updater is disabled. This policy does not control package-manager or extension updates.
macOS configuration profiles
Configuration profiles manage settings on macOS devices. A profile is an XML file (.mobileconfig) with key/value pairs that correspond to available policies.
These profiles can be deployed using Mobile Device Management (MDM) solutions or installed manually on individual devices.
Step 1: Obtain the sample configuration profile
Starting from VS Code version 1.99, each release ships with a sample .mobileconfig file. Follow these steps to locate the sample file on a macOS device with VS Code installed:
- Open Finder and navigate to
/Applications. - Right-click on Visual Studio Code.app (or your VS Code variant) and select Show Package Contents.
- Navigate to
Contents/Resources/app/policies. - Locate the file with the
.mobileconfigextension.
Step 2: Configure policy values
-
Copy the sample
.mobileconfigfile to a working location (for example, your Desktop or Documents folder). -
Open the copied file in a text editor (for example, TextEdit, VS Code, or any XML editor).
-
Find the application payload dictionary inside the
PayloadContentarray. Replace the corresponding policy entries with the values you want to enforce. For example:<key>AllowedExtensions</key> <string>{"microsoft": true, "github": true}</string> <key>UpdateMode</key> <string>start</string> <key>EnableFeedback</key> <true/> <key>TelemetryLevel</key> <string>off</string>This is a fragment of the application payload, not a complete configuration profile. Keep the profile's metadata and enclosing XML structure. Object-valued policies such as
AllowedExtensionsuse JSON encoded inside an XML<string>element. -
Remove the
<key>element and its associated value for each policy you don't want to enforce. Don't add duplicate entries for a policy. -
Save the
.mobileconfigfile.
Check both the profile's XML syntax and any JSON-encoded policy values before deployment. If a value isn't applied, run Developer: Policy Diagnostics and check the Window log for errors by running Developer: Show Logs and selecting Window.
Step 3: Deploy the policies
You can now deploy the configured policies at scale to all relevant devices in your organization with an MDM solution. You can manually test the policies on a local machine before deploying them at scale.
Deploy profiles at scale
For enterprise deployments across multiple devices, use Mobile Device Management (MDM) solutions such as:
- Microsoft Intune
- Apple Business Manager with MDM
For installation and removal steps for your macOS version, refer to Apple's configuration profile documentation.
Manually test policies on a local machine
Configure policies manually
Follow these steps to manually test your VS Code policy configuration on a macOS device before deploying at scale:
Step 1: Install the configuration profile
- Save your edited
.mobileconfigfile. - Double-click the
.mobileconfigfile in Finder. - Open System Settings > General > Device Management. On older macOS versions, find Profiles in System Settings or System Preferences.
- Open the downloaded profile, review its details, and follow the prompts to install it.
- If prompted, authenticate with your administrator credentials.
- Confirm the installation when prompted.
Step 2: Verify the profile installation
- Open System Settings > General > Device Management, or Profiles on older macOS versions.
- Verify that your VS Code configuration profile appears in the list.
- Restart VS Code and follow Verify policy enforcement. For the example above, confirm that
TelemetryLevelisoffandUpdateModeisstart.
Policies take effect immediately for new VS Code instances. You may need to restart VS Code if it's already running.
Remove a configuration profile
To remove a manually installed test profile:
- Open System Settings > General > Device Management, or Profiles on older macOS versions.
- Select the VS Code configuration profile.
- Select the Remove (or -) button.
- Authenticate with your administrator credentials to confirm removal.
For profiles deployed through device management, remove the assignment through that management solution. Removing a profile stops it from enforcing its values; other policies and user settings can still apply.
Linux JSON policies
Starting from VS Code version 1.106, you can configure VS Code setting policies on Linux devices by placing a JSON policy file at /etc/vscode/policy.json. This approach uses a simple JSON format to define policy values.
Deploy the file through your configuration management solution or install it manually on a test device.
Step 1: Obtain the sample policy file
Starting from VS Code version 1.106, each release ships with a sample policy.json file. You can obtain it from either an existing installation or by downloading and extracting the VS Code archive. The file is located in the resources/app/policies directory.
Step 2: Configure policy values
-
Copy the sample
policy.jsonfile to a working location. For a package installation under/usr/share/code, run:sudo cp /usr/share/code/resources/app/policies/policy.json /tmp/policy.json -
Edit the file using your preferred text editor:
sudo nano /tmp/policy.json # or sudo vim /tmp/policy.json # or code /tmp/policy.json -
Edit the policy values according to your requirements:
Example
policy.jsonconfiguration:{ "AllowedExtensions": { "microsoft": true, "github": true }, "UpdateMode": "start", "EnableFeedback": true, "TelemetryLevel": "off" }ImportantIf there's a syntax error in the policy value, the setting will not be applied. Run Developer: Show Logs from the Command Palette (⇧⌘P (Windows, Linux Ctrl+Shift+P)), select Window, and check for errors.
-
Remove the corresponding JSON property for each policy you don't want to enforce.
Use policy names such as TelemetryLevel, not editor setting IDs such as telemetry.telemetryLevel, as the top-level keys.
Step 3: Deploy the policies
You can now deploy the configured policies at scale to all relevant devices in your organization with an MDM solution. You can manually test the policies on a local machine before deploying them at scale.
Deploy policies at scale
For enterprise Linux deployments across multiple devices, use configuration management tools such as Ansible, Puppet, Chef, or Salt to deploy the policy.json file.
These tools allow administrators to deploy, update, and remove policies remotely across all managed Linux devices in the organization.
Manually test policies on a local machine
Step 1: Copy the policy file
-
Ensure the
/etc/vscodedirectory exists:sudo mkdir -p /etc/vscodeNoteYou need root or sudo privileges to create the directory and manage policy files in
/etc/vscode. -
Copy the edited policy file to the
/etc/vscode/system location:sudo cp /tmp/policy.json /etc/vscode/policy.jsonSet appropriate permissions:
sudo chmod 644 /etc/vscode/policy.json sudo chown root:root /etc/vscode/policy.json
Step 2: Verify the policy installation
- Launch VS Code (or restart it if already running).
- Follow Verify policy enforcement. For the example above, confirm that
TelemetryLevelisoffandUpdateModeisstart. - Open Settings (⌘, (Windows, Linux Ctrl+,)) and check that the corresponding settings are managed by your organization.
You can verify the policy file is being read by checking VS Code's logs or by attempting to change a managed setting (the change will be prevented).
Remove policies
To remove the policies delivered through this file, delete /etc/vscode/policy.json and restart VS Code. Other policy sources and user settings can still apply.
Verify policy enforcement
After you deploy enterprise policies to a device, use Developer: Policy Diagnostics to inspect the policies received by VS Code and their effective values. The command opens a new untitled Markdown document with a report of the current policy state on the device. It works the same way on Windows, macOS, and Linux.
The report includes the following sections:
- System Information: VS Code product name, version, and commit, useful for matching the report to a specific build.
- Account Information: details of the default account that is signed in, including the raw account-level policy data returned by the account provider.
- Account Policy Gate: state of the approved GitHub organizations gate that controls AI features. Possible states are
inactive,satisfied, andrestricted. When the state isrestricted, the report also lists a reason such asnoAccount,wrongProvider,orgNotApproved, orpolicyNotResolved. - Policy-Controlled Settings: two tables that list the policy state for each registered setting:
- Applied Policy: settings that are currently overridden by a policy, with the setting key, policy name, policy source, default value, current value, and the value enforced by the policy.
- Non-applied Policy: registered policies that are not currently being enforced. Use this table to detect deployment errors, such as a misspelled key or a policy file that is not being read.
- Authentication Information: registered authentication providers, sessions, accounts, and the extensions that have access to each account.
The report can contain sensitive information such as account identifiers, session details, and the list of extensions with access to each account. Review the contents before you share the report.
If the Account Policy Gate state is restricted with the reason policyNotResolved, run the Developer: Sync Account Policy command to force a refresh of the account-side policy data, then regenerate the report.
If a policy is missing or has an unexpected value, check delivery and precedence. If the reported value is correct but the behavior differs, check enforcement scope. The report alone does not prove that a restriction applies to every workflow or agent harness.
VS Code enterprise policy reference
The following table lists all available enterprise policies in VS Code.
For deployment values and examples, use the extension, AI, telemetry, and update guides. AI policy behavior can differ by harness; review the device-policy limitations before relying on a control.
| Policy Setting ID |
Description |
|---|---|
AgentsVoiceagents.voice.enabled |
Enable the Voice Mode panel in the chat view for voice-driven coding conversations. |
ChatAgentModechat.agent.enabled |
When enabled, agent mode can be activated from chat and tools in agentic contexts with side effects can be used. |
ChatAgentNetworkFilterchat.agent.networkFilter |
When enabled, network access by agent tools (fetch tool, integrated browser) is restricted according to #chat.agent.sandbox.network.allowedDomains# and #chat.agent.sandbox.network.deniedDomains#. Domain filtering is also applied to those tools when #chat.agent.sandbox.enabled# is enabled. Changes may not take full effect until VS Code is restarted. |
ChatAgentSandboxAllowUnsandboxedCommandschat.agent.sandbox.allowUnsandboxedCommands |
Deprecated. Controls whether Local agent terminal commands can run outside the sandbox after user confirmation when sandbox restrictions would block the command. Use the sandbox.allowBypass managed setting to control sandbox bypass in Copilot Agent Host. |
ChatAgentSandboxEnabledchat.agent.sandbox.enabled |
Deprecated. Controls whether the Local agent uses sandboxing to restrict what tools can do. In Copilot Agent Host, this legacy policy supplies an overridable sandbox default, not a mandatory requirement. Use the runtime-native sandbox.enabled managed setting to require sandboxing in Copilot Agent Host. |
ChatAgentAllowedNetworkDomainschat.agent.sandbox.network.allowedDomains |
Allowed domains for network access by agent tools (fetch tool, integrated browser) when #chat.agent.networkFilter# is enabled. In Copilot Agent Host sessions with sandboxing (#chat.agent.sandbox.enabled#) enabled, also restricts the integrated browser when outbound network access (#chat.agent.sandbox.network.allowNetwork#) is allowed; an empty allow list adds no sandbox restriction. This list does not enable outbound access. Supports wildcards like *.example.com. Outside sandboxing, when both allowed and denied lists are empty, all domains are blocked. Denied domains (see #chat.agent.sandbox.network.deniedDomains#) take precedence. Changes may not take full effect until VS Code is restarted. |
ChatAgentSandboxAllowNetworkchat.agent.sandbox.network.allowNetwork |
Deprecated. Controls whether the Local agent sandbox allows outbound connections to all network domains while preserving file system restrictions. Use the sandbox.userPolicy.network.allowOutbound managed setting to enforce outbound network restrictions in Copilot Agent Host. |
ChatAgentDeniedNetworkDomainschat.agent.sandbox.network.deniedDomains |
Denied domains for network access by agent tools (fetch tool, integrated browser) when #chat.agent.networkFilter# is enabled. Also applies to the integrated browser in sandboxed Copilot Agent Host sessions (#chat.agent.sandbox.enabled#); allowing outbound network access (#chat.agent.sandbox.network.allowNetwork#) does not bypass this list. Takes precedence over #chat.agent.sandbox.network.allowedDomains#. Supports wildcards like *.example.com. Changes may not take full effect until VS Code is restarted. |
Claude3PIntegrationchat.agentHost.claudeAgent.enabled github.copilot.chat.claudeAgent.enabled |
Enable Claude Agent sessions in VS Code. Start and resume agentic coding sessions powered by Anthropic Claude Agent SDK directly in the editor. Uses your existing Copilot subscription. |
Codex3PIntegrationchat.agentHost.codexAgent.enabled |
Enable Codex Agent sessions in VS Code. Start and resume agentic coding sessions powered by OpenAI Codex. Usage can be routed through GitHub Copilot or authenticated directly with an OpenAI account. |
ChatAgentHostCustomTerminalToolchat.agentHost.customTerminalTool.enabled |
Enable the Agent Host custom terminal tool override for Copilot SDK sessions. |
ChatAgentHostGitHubEnvironmentschat.agentHost.githubEnvironments.enabled |
Controls whether this machine can be shared as a GitHub environment and connect to other shared machines through Mission Control. |
CopilotOtelCaptureContentchat.agentHost.otel.captureContent github.copilot.chat.otel.captureContent |
Controls whether Copilot OpenTelemetry export captures prompt, response, and tool content. |
CopilotOtelCaptureIdentitygithub.copilot.chat.otel.captureIdentity |
Controls whether Copilot OpenTelemetry captures the authenticated account name, operating system username, and machine hostname. Independent of content capture. |
CopilotOtelEnabledchat.agentHost.otel.enabled github.copilot.chat.otel.enabled |
Controls whether Copilot OpenTelemetry export is enabled. When managed, users cannot override the enterprise value. |
CopilotOtelProtocolchat.agentHost.otel.exporterType github.copilot.chat.otel.exporterType |
Controls the enterprise-managed OTLP protocol for Copilot OpenTelemetry export. |
CopilotOtelHeadersgithub.copilot.chat.otel.headers |
Controls the enterprise-managed OTLP exporter headers for Copilot OpenTelemetry export. |
CopilotOtelEndpointchat.agentHost.otel.otlpEndpoint github.copilot.chat.otel.otlpEndpoint |
Controls the enterprise-managed OTLP collector endpoint for Copilot OpenTelemetry export. |
CopilotOtelOtlpProtocolgithub.copilot.chat.otel.protocol |
Controls the enterprise-managed OTLP wire protocol (protobuf vs JSON) for Copilot OpenTelemetry export. |
CopilotOtelOutfilechat.agentHost.otel.outfile github.copilot.chat.otel.outfile |
Controls the enterprise-managed file export destination for Copilot OpenTelemetry and prevents local file export when an enterprise OTLP destination is configured. |
CopilotOtelResourceAttributesgithub.copilot.chat.otel.resourceAttributes |
Controls the enterprise-managed OTel resource attributes for Copilot OpenTelemetry export. |
CopilotOtelServiceNamegithub.copilot.chat.otel.serviceName |
Controls the enterprise-managed OTel service.name resource attribute for Copilot OpenTelemetry export. |
ChatApprovedAccountOrganizations |
Setting this policy to a non-empty list activates the Approved Account gate: all AI features are disabled until the user signs into a GitHub account whose organizations intersect this list AND the account-side policy data has resolved. Comparison is case-insensitive. Use '*' as a wildcard to accept any signed-in GitHub or GHE account (use this for GHE deployments where the organization list is not surfaced). |
ChatCanvaseschat.canvases.enabled |
Controls whether Copilot sessions can use Canvases to present interactive content. |
CopilotCliCommandchat.copilotCliCommand.enabled |
Controls whether VS Code provides the copilot command for GitHub Copilot CLI. When disabled, VS Code setup on Windows and the Linux DEB and RPM packages don't add the command to PATH and remove the entry they added, VS Code doesn't install GitHub Copilot CLI, and integrated terminals don't include the command. A GitHub Copilot CLI that was installed separately isn't affected. |
ChatCopilotConnectorschat.customizations.copilotConnectors.enabled |
Enables Copilot connectors in the unified marketplace and exposes connected connector MCP servers in Chat Customizations and Copilot agent sessions. |
ChatCustomizationMarketplacechat.customizations.marketplace.enabled |
Shows Browse instead of Overview when a customization marketplace source is enabled. When disabled, marketplace discovery remains in the existing customization management pages. |
ChatStrictPluginOnlyCustomization |
Blocks standalone user and workspace skills, agents, hooks, instructions, and MCP servers while keeping eligible plugin customizations available. When delivered through Copilot managed settings, also blocks executable Copilot extensions from user, workspace, and session directories. |
ChatDefaultConfigurationchat.defaultConfiguration |
Sets the default configuration for new agent sessions, as a JSON object with an optional "mode" ("interactive", "plan", or "autopilot") and "approvals" ("manual", "assisted", or "allowAll"). Users can still change them for each session. |
ChatDefaultModelchat.defaultModel |
Sets the default chat model for new conversations. Accepts "auto", a model family name (such as "opus" or "gemini"), or a full model id. Users can still switch the model within a conversation. |
ChatEditorPreferCopilotHarnesschat.editor.preferCopilotHarness |
Configure whether VS Code uses the Agent Host Copilot SDK instead of the local harness for new editor chat sessions. |
ChatAgentExtensionToolschat.extensionTools.enabled |
Enable using tools contributed by third-party extensions. |
ChatAllowManagedHooksOnly |
Allows hooks only from enterprise-managed sources and plugins force-enabled by policy. |
ChatMCPchat.mcp.access |
Controls access to installed Model Context Protocol servers. |
ChatAllowedMcpServers |
Allowlist of Model Context Protocol servers. When set, only servers matching an entry may be installed or run; omit entirely to allow all servers (subject to the deny list). |
ChatAllowManagedMcpServersOnly |
Use only the enterprise-managed MCP allowlist when deciding which servers may run. |
ChatDeniedMcpServers |
Denylist of Model Context Protocol servers. Servers matching any entry are blocked from being installed or run, even if they also match the allow list; deny rules always take precedence. |
McpGalleryServiceUrl |
Configure the MCP Gallery service URL to connect to |
ChatPluginsEnabledchat.plugins.enabled |
Enable agent plugin integration in chat. |
ChatEnabledPluginschat.plugins.enabledPlugins |
Plugin enablement. Keys are plugin IDs in {plugin}@{marketplace} form; values enable or disable the plugin. |
ChatExtraMarketplaces |
Additional plugin marketplaces to query. Keys are marketplace names; values are GitHub shorthand (owner/repo[#ref]) or Git URIs ({url}[#ref]), optionally with an enterprise-managed auto-update override. |
CopilotSessionSyncchat.sessionSync.enabled |
Enable session sync to GitHub.com for cross-device Copilot session history. When disabled by organization policy, session data is kept local only. |
ChatTemporaryDefaultToLocalHarness |
Temporarily keeps Local available and prevents automatic selection of the Copilot harness, overriding rollout settings and managed sandbox and permission routing. Local does not enforce runtime-only sandbox and permission settings. |
ChatToolsEligibleForAutoApprovalchat.tools.eligibleForAutoApproval |
Controls which tools are eligible for automatic approval. Tools set to 'false' will always present a confirmation and will never offer the option to auto-approve. The default behavior (or setting a tool to 'true') may result in the tool offering auto-approval options. |
ChatToolsAutoApprovechat.tools.global.autoApprove |
Global auto approve also known as "YOLO mode" disables manual approval completely for all tools in all workspaces, allowing the agent to act fully autonomously. This is extremely dangerous and is never recommended, even containerized environments like Codespaces and Dev Containers have user keys forwarded into the container that could be compromised. This feature disables critical security protections and makes it much easier for an attacker to compromise the machine. Note: This setting only controls tool approval and does not prevent the agent from asking questions. To automatically answer agent questions, use the #chat.autoReply# setting. |
ChatToolsTerminalEnableAutoApprovechat.tools.terminal.enableAutoApprove |
Controls whether to allow auto approval in the run in terminal tool. |
ChatHookschat.useHooks |
Controls whether chat hooks are executed at strategic points during an agent's workflow. Hooks are loaded from the files configured in #chat.hookFilesLocations#. This setting is only used by the Local agent harness. |
DictationEnableddictation.enabled |
Controls whether dictation is available across the product (chat input, editor, and terminal). |
DictationLLMCleanupdictation.experimental.llmCleanup |
Controls whether final dictation transcripts are sent to a language model for cleanup. |
DictationModeldictation.model |
Controls the transcription model used for dictation. |
AllowedExtensionsextensions.allowed |
Specify a list of extensions that are allowed to use. This helps maintain a secure and consistent development environment by restricting the use of unauthorized extensions. More information: https://aka.ms/vscode/enterprise/extensions/allowed |
ExtensionsAutoUpdateextensions.autoUpdate |
Controls the automatic update behavior of extensions. The updates are fetched from a Microsoft online service. |
ExtensionsAutoUpdateDelayextensions.autoUpdateDelay |
Controls the delay in hours after an extension update is published before it is automatically installed. Only applies when #extensions.autoUpdate# is set to on. This delay helps avoid installing potentially problematic updates immediately after release. |
ExtensionGalleryAuthProvider |
Configure the authentication provider for the Extensions Marketplace |
ExtensionGalleryServiceUrl |
Configure the Marketplace service URL to connect to |
CopilotReviewAgentgithub.copilot.chat.reviewAgent.enabled |
Enables the code review agent. |
CopilotReviewSelectiongithub.copilot.chat.reviewSelection.enabled |
Enables code review on current selection. |
CopilotNextEditSuggestionsgithub.copilot.nextEditSuggestions.enabled |
Whether to enable next edit suggestions (NES). NES can propose a next edit based on your recent changes. |
McpEnterpriseManagedAuthIdp |
The OAuth/OIDC IdP configuration used for enterprise-managed Model Context Protocol (MCP) server authentication. |
EnableFeedbacktelemetry.feedback.enabled |
Enable feedback mechanisms such as the issue reporter, surveys, and other feedback options. |
TelemetryLeveltelemetry.telemetryLevel |
Controls the level of telemetry. |
UpdateModeupdate.mode |
Configure whether you receive automatic updates. The updates are fetched from a Microsoft online service. |
UpdateRestartWhenIdleupdate.restartWhenIdle |
Controls whether to restart automatically to apply a downloaded update after the screen has been locked for a minute while on power, keeping the machine from going to sleep until then. The restart only happens if every window can be restored exactly as it was, for example when no terminal process, task, debug session, chat request, or agent session is running, and it does not take focus from another application. Only supported on macOS. |
BrowserChatToolsworkbench.browser.enableChatTools |
When enabled, chat agents can use browser tools to open and interact with pages in the Integrated Browser. |
If you want to enact more policies, open an issue in the VS Code GitHub repository. The team will determine if there is already a corresponding setting for the behavior or if a new setting and policy should be created.