One read of Claude Documentationclaude-docs-20261023T170712Z
19 pages moved out of 280 read.
Pages moved
19
significant first
Pages read
280
in this capture
Captured
17:07 UTC
Corpus hash
3b46dd2b6b7f
corpus-hash
What this read moved
1-19 of 19government/desktop/import Changed · +7 / -7 lines
from line 8
88
99## Before you begin
1010
11<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/gGFKuNSbKYs4JMmK/images/government/videos/user-01-signing-in-for-the-first-time.mp4?fit=max&auto=format&n=gGFKuNSbKYs4JMmK&q=85&s=fcfdbcaa5f3ee9e9496e6014ed516b65" aria-label="Video walkthrough: Signing in for the first time" />
11<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/8qHdPdet0sX9-CKy/images/government/videos/user-01-signing-in-for-the-first-time-v2.mp4?fit=max&auto=format&n=8qHdPdet0sX9-CKy&q=85&s=4383be097080bd0b4e0b73ef155afdef" aria-label="Video walkthrough: Signing in for the first time" />
1212
1313Video: Signing in for the first time (1 min 32 s). Narrated with an AI-generated voice, with on-screen captions.
1414
from line 32
3232>
3333> Code is for software work, in a folder you choose on your computer.
3434>
35> If you can sign in but the model picker is empty, you have no seat yet. Ask your organization's owner to assign you one.
35> If you can sign in but the model picker is empty, ask your organization's owner to check your seat tier and its models.
3636
3737* **You have an account on the web app.** It must use the same work email address as your account in Claude Desktop.
3838* **You are signed in to the web app in your default browser.** The import opens a browser tab there and asks for a one-time code, which expires after a few minutes.
from line 41
4141
4242## Run the import
4343
44<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/gGFKuNSbKYs4JMmK/images/government/videos/user-02-bringing-your-chats-and-projects-over.mp4?fit=max&auto=format&n=gGFKuNSbKYs4JMmK&q=85&s=b7860d7dd7807b95aecc486b1aaf26f7" aria-label="Video walkthrough: Bringing your chats and projects over from Claude for Government Web" />
44<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/8qHdPdet0sX9-CKy/images/government/videos/user-02-bringing-your-chats-and-projects-over-v2.mp4?fit=max&auto=format&n=8qHdPdet0sX9-CKy&q=85&s=17d612064ed90a2edf86e70bcf4bf72e" aria-label="Video walkthrough: Bringing your chats and projects over from Claude for Government Web" />
4545
46Video: Bringing your chats and projects over from Claude for Government Web (2 min 10 s). Narrated with an AI-generated voice, with on-screen captions.
46Video: Bringing your chats and projects over from Claude for Government Web (2 min 12 s). Narrated with an AI-generated voice, with on-screen captions.
4747
4848> **Transcript**
4949>
50> If you used Claude for Government - Web, you can copy your chats and projects into Claude Desktop. First, sign in to the web app in your browser with the same work email you use in Claude Desktop.
50> If you used Claude for Government Web, you can copy your chats and projects into Claude Desktop. First, sign in to the web app in your browser with the same work email you use in Claude Desktop.
5151>
52> In Claude Desktop, open Settings from the account menu at the bottom of the sidebar, then Import and export, and select Import. The app does not prompt you to do this. Run it when you are ready.
52> In Claude Desktop, open Settings from the account menu at the bottom of the sidebar, then Import and export, and select Import. Run it when you are ready.
5353>
5454> Select Sign in to Claude for Government Web, then Sign in. Claude shows a one-time code and opens the web app in your browser.
5555>
from line 65
6565>
6666> A project that had instructions shows a notice. Claude won't follow imported instructions until you accept them: select Review instructions, check them, and save, or choose Use as is.
6767>
68> The import only copies. Your chats and projects stay in the web app too. You can run it again, and chats already imported are skipped. If the Import page says import isn't enabled, ask your administrator.
68> The import only copies. Your chats and projects stay in the web app too. You can run it again, and chats already imported are skipped. If the Import and export page says import isn't enabled, update Claude Desktop. If it still says so, ask your administrator.
6969
7070You start the import yourself from **Settings**, whenever you are ready. If Claude Desktop's home screen shows a **Pick up where you left off in Claude for Government Web** banner, its **Start import** button takes you to the **Import & export** page in **Settings**. If you click **Skip for now** or close the banner, it can come back a few days later, until you run the import or have skipped it a few times. Claude Desktop versions earlier than 2.16120.0 hide the banner permanently after one skip.
7171
government/desktop/video-walkthroughs Changed · +13 / -13 lines
from line 14
1414
1515Signing in to Claude Desktop with your work account and a pairing code, and where Chat, Cowork, Code, and the model picker are.
1616
17<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/gGFKuNSbKYs4JMmK/images/government/videos/user-01-signing-in-for-the-first-time.mp4?fit=max&auto=format&n=gGFKuNSbKYs4JMmK&q=85&s=fcfdbcaa5f3ee9e9496e6014ed516b65" aria-label="Video walkthrough: Signing in for the first time" />
17<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/8qHdPdet0sX9-CKy/images/government/videos/user-01-signing-in-for-the-first-time-v2.mp4?fit=max&auto=format&n=8qHdPdet0sX9-CKy&q=85&s=4383be097080bd0b4e0b73ef155afdef" aria-label="Video walkthrough: Signing in for the first time" />
1818
1919Video: Signing in for the first time (1 min 32 s). Narrated with an AI-generated voice, with on-screen captions.
2020
from line 38
3838>
3939> Code is for software work, in a folder you choose on your computer.
4040>
41> If you can sign in but the model picker is empty, you have no seat yet. Ask your organization's owner to assign you one.
41> If you can sign in but the model picker is empty, ask your organization's owner to check your seat tier and its models.
4242
4343### Bringing your chats and projects over from Claude for Government Web
4444
4545Copying your conversations and projects from Claude for Government Web into Claude Desktop. Read more: [Run the import](/docs/government/desktop/import.md#run-the-import).
4646
47<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/gGFKuNSbKYs4JMmK/images/government/videos/user-02-bringing-your-chats-and-projects-over.mp4?fit=max&auto=format&n=gGFKuNSbKYs4JMmK&q=85&s=b7860d7dd7807b95aecc486b1aaf26f7" aria-label="Video walkthrough: Bringing your chats and projects over from Claude for Government Web" />
47<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/8qHdPdet0sX9-CKy/images/government/videos/user-02-bringing-your-chats-and-projects-over-v2.mp4?fit=max&auto=format&n=8qHdPdet0sX9-CKy&q=85&s=17d612064ed90a2edf86e70bcf4bf72e" aria-label="Video walkthrough: Bringing your chats and projects over from Claude for Government Web" />
4848
49Video: Bringing your chats and projects over from Claude for Government Web (2 min 10 s). Narrated with an AI-generated voice, with on-screen captions.
49Video: Bringing your chats and projects over from Claude for Government Web (2 min 12 s). Narrated with an AI-generated voice, with on-screen captions.
5050
5151> **Transcript**
5252>
53> If you used Claude for Government - Web, you can copy your chats and projects into Claude Desktop. First, sign in to the web app in your browser with the same work email you use in Claude Desktop.
53> If you used Claude for Government Web, you can copy your chats and projects into Claude Desktop. First, sign in to the web app in your browser with the same work email you use in Claude Desktop.
5454>
55> In Claude Desktop, open Settings from the account menu at the bottom of the sidebar, then Import and export, and select Import. The app does not prompt you to do this. Run it when you are ready.
55> In Claude Desktop, open Settings from the account menu at the bottom of the sidebar, then Import and export, and select Import. Run it when you are ready.
5656>
5757> Select Sign in to Claude for Government Web, then Sign in. Claude shows a one-time code and opens the web app in your browser.
5858>
from line 68
6868>
6969> A project that had instructions shows a notice. Claude won't follow imported instructions until you accept them: select Review instructions, check them, and save, or choose Use as is.
7070>
71> The import only copies. Your chats and projects stay in the web app too. You can run it again, and chats already imported are skipped. If the Import page says import isn't enabled, ask your administrator.
71> The import only copies. Your chats and projects stay in the web app too. You can run it again, and chats already imported are skipped. If the Import and export page says import isn't enabled, update Claude Desktop. If it still says so, ask your administrator.
7272
7373### Chat, Cowork, and Code
7474
75What Chat, Cowork, and Code are each for, when Claude asks for your approval in each, and how projects work.
75What Chat, Cowork, and Code are each for, when Claude asks for your approval, and how projects work.
7676
77<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/gGFKuNSbKYs4JMmK/images/government/videos/user-03-chat-cowork-and-code.mp4?fit=max&auto=format&n=gGFKuNSbKYs4JMmK&q=85&s=9402b32239886fb0f79a8648921fe53e" aria-label="Video walkthrough: Chat, Cowork, and Code: which one to use" />
77<video controls preload="metadata" playsInline src="https://mintcdn.com/claude-ai/8qHdPdet0sX9-CKy/images/government/videos/user-03-chat-cowork-and-code-v2.mp4?fit=max&auto=format&n=8qHdPdet0sX9-CKy&q=85&s=da546543ca5b688980a79a33e1a41e22" aria-label="Video walkthrough: Chat, Cowork, and Code: which one to use" />
7878
79Video: Chat, Cowork, and Code: which one to use (2 min 35 s). Narrated with an AI-generated voice, with on-screen captions.
79Video: Chat, Cowork, and Code: which one to use (2 min 26 s). Narrated with an AI-generated voice, with on-screen captions.
8080
8181> **Transcript**
8282>
from line 88
8888>
8989> Then Claude searches and answers.
9090>
91> Chat also reads files you attach. Add a document with the plus button and ask for a summary. Out of the box, Chat doesn't reach into folders on your computer.
91> Chat also reads files you attach. Add a document with the plus button and ask for a summary.
9292>
9393> For work that should end up in your files, switch to Cowork and choose a folder. Claude asks whether it may change files there. Select Allow.
9494>
from line 96
9696>
9797> When it finishes, the summary is a real file in your folder.
9898>
99> Code is Claude Code, built into the app, for software development. It works directly in the folder you open, and here it asks before running a command.
99> Code is Claude Code, built into the app, for software development. It works directly in the folder you open.
100100>
101101> Projects keep ongoing work in one place. Under Projects, select New project, then Use an existing folder. Choose the folder, name the project, and create it.
102102>
103> A project holds your folder and instructions, and you start chats or Cowork tasks in it. Projects live on your computer, not in the cloud. They are personal to you and, unlike on Claude for Government - Web, cannot be shared with colleagues.
103> A project holds your folder and instructions, and you start chats or Cowork tasks in it. Projects live on your computer, not in the cloud. They are personal to you and, unlike on Claude for Government Web, cannot be shared with colleagues.
104104>
105105> Chat for questions and documents. Cowork for tasks in your folders. Code for software development. Your organization decides which of the three you have, so a colleague may see a different set. If one is missing for you, ask your organization's owner.
106106
third-party/claude-desktop/code Changed · +3 / -2 lines
from line 43
4343| `coworkEgressAllowedHosts` | A network sandbox restricted to your hosts, `allowManagedDomainsOnly`, and `WebFetch` allow rules for the same hosts, which [don't block other hosts](/docs/third-party/claude-desktop/web-tools.md#how-the-egress-allowlist-applies-in-code-sessions). Claude Code's own inference and telemetry connections run outside the sandbox. A shell command or [HTTP hook](https://code.claude.com/docs/en/hooks#http-hook-fields) that calls your inference provider, `api.anthropic.com`, or your OTLP collector needs that host in the list. |
4444| `allowedWorkspaceFolders` | A filesystem sandbox for shell commands, which can then create or change files only inside your allowed roots and the session's temporary locations. The sandbox does not restrict which files those commands read unless you also set `blockReadsOutsideWorkingDirectories`. The roots are also passed as `additionalDirectories` at launch, which is always applied independent of managed-settings precedence, and the app keeps Claude's file tools inside the roots. The app also refuses to start a Code session outside an allowed root. |
4545| `blockReadsOutsideWorkingDirectories` | Claude Code's `permissions.blockReadsOutsideWorkingDirectories` for Code sessions. Claude's file tools refuse to read outside the working directories (the session's folder plus your allowed roots, if any) in every permission mode. Where the sandbox from `allowedWorkspaceFolders` or `coworkEgressAllowedHosts` is running, it also hides the user's home directory and similar locations, such as other users' home folders and mounted volumes, from shell commands, so a sandboxed command that reads there fails without a prompt. Without a running sandbox, a shell command that reads outside the working directories, or that Claude Code cannot analyze, asks the user for approval first, even in bypass permissions mode. The key requires Claude Desktop 1.46388.1 or later. The block takes effect only in sessions that run Claude Code v2.1.257 or later; if the app cannot install its current Claude Code engine and a session runs one older than v2.1.257 that is still on the device, that session runs without the block and the app logs a warning. |
46| `managedMcpServers` | `strictPluginOnlyCustomization` set to `["mcp"]`, so the Code session does not load MCP servers that users define on Claude Code's side (`~/.claude.json`, a project's `.mcp.json`, or `claude mcp add`); your managed servers, which the app connects and supplies to the session itself, and MCP servers bundled in plugins still load. When [`isLocalDevMcpEnabled`](/docs/third-party/claude-desktop/configuration#islocaldevmcpenabled) is `false`, the app also sets an `allowedMcpServers` list that admits only remote servers, with `allowManagedMcpServersOnly`, so local (stdio) servers bundled in plugins from marketplaces or that users install themselves are refused, while those plugins' remote servers still connect. Per-tool `toolPolicy` values on each server are emitted as `permissions.deny` (for `blocked`) or `permissions.ask` (for `ask`) rules against the corresponding `mcp__<server>__<tool>` names. |
46| `managedMcpServers` | `strictPluginOnlyCustomization` set to `["mcp"]`, so the Code session does not load MCP servers that users define on Claude Code's side (`~/.claude.json`, a project's `.mcp.json`, or `claude mcp add`); your managed servers, which the app connects and supplies to the session itself, and MCP servers bundled in plugins still load. Per-tool `toolPolicy` values on each server are emitted as `permissions.deny` (for `blocked`) or `permissions.ask` (for `ask`) rules against the corresponding `mcp__<server>__<tool>` names. |
47| `isLocalDevMcpEnabled` | When `false` and `allowedPluginMcpServers` is unset, an `allowedMcpServers` list that admits only remote servers, with `allowManagedMcpServersOnly`, whether or not `managedMcpServers` is set. The Code session refuses local (stdio) servers bundled in plugins from marketplaces or that users install themselves, and local servers that users define on Claude Code's side (`~/.claude.json`, a project's `.mcp.json`, or `claude mcp add`). Those plugins' remote servers still connect. Your managed servers and the servers from the organization plugins directory, which the app connects and supplies to the session itself, aren't subject to the list. |
4748| `allowedPluginMcpServers` | `strictPluginOnlyCustomization` set to `["mcp"]` together with an `allowedMcpServers` list holding your entries and `allowManagedMcpServersOnly`, whether or not `managedMcpServers` is set. MCP servers bundled in plugins from marketplaces, or in plugins users install themselves, connect only when their URL matches an entry; no such plugin's local (stdio) server is admitted, and an empty list admits none. Your managed servers and the servers from the organization plugins directory still load. |
4849| `deniedPluginMcpServers` | The app sets a `deniedMcpServers` list holding your entries, whether or not `allowedPluginMcpServers` is set. When the key is unset or its list is empty, the app sets no `deniedMcpServers` list. When the value is not a list, or holds an entry that the app can't read, the app sets a list whose one entry is `*`, a pattern that matches every URL. See the [`deniedPluginMcpServers` reference](/docs/third-party/claude-desktop/configuration#deniedpluginmcpservers) for what the setting does and what an entry matches. |
4950| `userPluginMarketplacesEnabled` | When `false`, a `strictKnownMarketplaces` allowlist. See [Controlling user extensions](/docs/third-party/claude-desktop/extensions.md#controlling-user-extensions) for what the key does. |
5051| `builtinToolPolicy`, argument-scoped rules such as `Bash(curl *)` | A `permissions.ask` rule for each rule set to `"ask"`. |
5152
52The network and filesystem sandboxes apply on macOS, and on Linux devices and [SSH hosts](/docs/third-party/claude-desktop/ssh-remote-sessions.md#managed-configuration-on-the-remote-host) with Claude Code's [sandbox dependencies](https://code.claude.com/docs/en/sandboxing) installed. Claude Code doesn't sandbox shell commands on Windows devices, and on a Linux device or SSH host without the dependencies commands run unsandboxed. In those cases, and when `coworkEgressAllowedHosts` contains `*` and `allowedWorkspaceFolders` is unset, `blockReadsOutsideWorkingDirectories` still confines Claude's file tools but can only ask the user to approve shell commands that read outside the working directories or that Claude Code can't verify. For what a remote session shows, see [Sandbox status on the remote host](/docs/third-party/claude-desktop/ssh-remote-sessions.md#sandbox-status-on-the-remote-host).
53The network and filesystem sandboxes apply on macOS, and on Linux devices and [SSH hosts](/docs/third-party/claude-desktop/ssh-remote-sessions.md#managed-configuration-on-the-remote-host) with Claude Code's [sandbox dependencies](https://code.claude.com/docs/en/sandboxing) installed. Claude Code doesn't sandbox shell commands on Windows devices, and on a Linux device or SSH host without the dependencies commands run unsandboxed. When `coworkEgressAllowedHosts` contains `*` and `allowedWorkspaceFolders` is unset, Claude Desktop turns on neither sandbox. In all of these cases, `blockReadsOutsideWorkingDirectories` still confines Claude's file tools but can only ask the user to approve shell commands that read outside the working directories or that Claude Code can't verify. For what a remote session shows, see [Sandbox status on the remote host](/docs/third-party/claude-desktop/ssh-remote-sessions.md#sandbox-status-on-the-remote-host).
5354
5455> [!WARNING]
5556>
third-party/claude-desktop/configuration Changed · +8 / -6 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 1231
12311231>
12321232> **Extended context** (`supports1m`) is a capability assertion you make about your deployment; only set it for models you've confirmed support the 1M-token window:
12331233>
1234> ```json theme={null}
1234> ```json theme={null} theme={null} theme={null}
12351235> [{"name": "claude-sonnet-5", "supports1m": true}, "claude-opus-4-8"]
12361236> ```
12371237>
from line 1239
12391239>
12401240> **Display label** (`labelOverride`) is for IDs the picker can't derive a friendly name from (Bedrock ARNs, gateway routing aliases). Display-only; `name` is still what the app sends:
12411241>
1242> ```json theme={null}
1242> ```json theme={null} theme={null} theme={null}
12431243> [{"name": "arn:aws:bedrock:us-east-1:123:application-inference-profile/abc", "labelOverride": "Claude Opus (Prod)"}]
12441244> ```
12451245>
12461246> **Tier mapping** (`anthropicFamilyTier`) tells the app which Claude tier (`haiku`/`sonnet`/`opus`/`fable`/`mythos`) an entry stands in for, so bare tier aliases (e.g. in Code sessions) resolve to your model. `isFamilyDefault: true` picks the winner when several entries share a tier:
12471247>
1248> ```json theme={null}
1248> ```json theme={null} theme={null} theme={null}
12491249> [{"name": "us.anthropic.claude-opus-4-8", "anthropicFamilyTier": "opus"}]
12501250> ```
12511251>
from line 1357
13571357>
13581358> Each row replaces Anthropic list price for one model in the Usage page's estimate, in USD per million tokens (`inputPerMtok`, `outputPerMtok`, `cacheReadPerMtok`, `cacheWritePerMtok`, all four required; `cacheWritePerMtok` prices both 5-minute and 1-hour cache writes); rows apply only while `inferenceModelPricingEnabled` is `true` and do not turn the estimate on by themselves. Mirrors Claude Code's managed `modelPricing.overrides`, and `name` is matched the same way: a built-in Claude model ID (e.g. `claude-sonnet-4-6`, or its Bedrock, Vertex, or Foundry ID) covers every dated and provider spelling of that model; any other value (a gateway alias, an inference-profile ARN) matches that exact ID only (case-insensitive) and wins over a built-in row. An ID Claude Code cannot map to a Claude model at all gets no estimate until a row here prices it. `inferenceModelPricingMultiplier` still applies on top of a row.
13591359>
1360> ```json theme={null}
1360> ```json theme={null} theme={null} theme={null}
13611361> {"inferenceModelPricingEnabled": true, "inferenceModelPricingMultiplier": 0.9, "inferenceModelPricing": [{"name": "claude-sonnet-4-6", "inputPerMtok": 2.4, "outputPerMtok": 12, "cacheReadPerMtok": 0.24, "cacheWritePerMtok": 3}]}
13621362> ```
13631363>
from line 3413
34133413
34143414**Hide feature announcements.** Suppress unprompted feature-announcement UI: the post-update “What’s new” nudge and new-feature tips. Users can still open release notes themselves.
34153415
3416[Self-hosted install](#self-hosted-column): Server side. Desktop hides its own announcements of new features.
3417
3418> **More detail**
3419>
3420> Covers the version-shipped announcement UI baked into each release: the
3416[Self-hosted install](#self-hosted-column): Server side. Desktop hides its own announcements of new f
third-party/claude-desktop/connectors-m365 Changed · +3 / -3 lines
from line 26
2626| Device egress | `login.microsoftonline.com` and the connector host | `login.microsoftonline.com`, `graph.microsoft.com`, and your tenant's SharePoint and OneDrive hosts |
2727| Device-based Conditional Access | Not supported (the server-side exchange has no device identity) | Supported on managed Windows and Mac devices through brokered sign-in |
2828| Write actions | Available with [write scopes on the Anthropic connector app](#control-write-actions-on-the-remote-connector) | Available with [write scopes](#grant-write-scopes) |
29| US Government clouds | Separate connector deployment; contact your Anthropic representative | Built in; set `azureCloud` |
29| US Government clouds | Separate connector deployment; contact your Anthropic representative | Built in; set `azureCloud` for GCC High and DoD |
3030
3131## Remote connector
3232
from line 288
288288| `server` | Yes | Must be `microsoft365`. Built-in entries use this field instead of `url`, `transport`, or `command`; an entry that mixes `server` with those fields is rejected. |
289289| `clientId` | Yes | The Application (client) ID of the local-mode app from step 1. |
290290| `tenantId` | Yes | Your Directory (tenant) ID. |
291| `azureCloud` | No | `global` (default), `us-gov-high`, or `us-gov-dod`. Selects the Microsoft Entra and Microsoft Graph hosts for US Government clouds. |
291| `azureCloud` | No | `global` (default), `us-gov-high`, or `us-gov-dod`. Selects the Microsoft Entra and Microsoft Graph hosts for GCC High and DoD. |
292292| `continuousAccessEvaluation` | No | `enabled` (default) or `disabled`. When enabled, the connector requests Continuous Access Evaluation-capable Microsoft Graph tokens, which live up to about 28 hours and stop working within minutes after an administrator revokes the user's sessions or disables the account in Entra, and, where your tenant enforces an IP named-location or Global Secure Access compliant-network Conditional Access policy, when the token is used from outside that network. `disabled` keeps standard one-hour tokens. A change applies to tokens issued after the connector next starts, and an already-issued token stays in use until it expires (select **Disconnect**, then **Connect**, to sign in again immediately). Requires Claude Desktop 1.49585.0 or later; earlier versions ignore the field and request standard one-hour tokens. |
293293| `scope` | No | Space-separated delegated Graph scopes to request instead of the default read set. The app rejects an entry that has a `scopes` array, so write the scopes as one string. See [Configure scopes](#configure-scopes). |
294294| `toolPolicy` | No | Per-tool approval locks, the same as for any managed server. See [`toolPolicy`](/docs/third-party/claude-desktop/configuration#managedmcpservers). |
from line 307
307307
308308To read Office documents, such as Word and PowerPoint files, the connector asks Microsoft Graph for a PDF version and follows the redirect Graph returns. Microsoft chooses the host in that redirect.
309309
310US Government cloud deployments use `login.microsoftonline.us` and `graph.microsoft.us` (or `dod-graph.microsoft.us` for `us-gov-dod`) instead of `login.microsoftonline.com` and `graph.microsoft.com`, matching the `azureCloud` setting. GCC High (`us-gov-high`) support has been confirmed in customer deployments. No egress to any Anthropic host is needed for Microsoft 365 data with the local connector.
310GCC High and DoD deployments use `login.microsoftonline.us` and `graph.microsoft.us` (or `dod-graph.microsoft.us` for `us-gov-dod`) instead of `login.microsoftonline.com` and `graph.microsoft.com`, matching the `azureCloud` setting. For a Microsoft 365 GCC tenant, leave `azureCloud` at `global`. GCC High (`us-gov-high`) support has been confirmed in customer deployments. No egress to any Anthropic host is needed for Microsoft 365 data with the local connector.
311311
312312### Configure scopes
313313
third-party/claude-desktop/extensions Changed · +10 / -9 lines
from line 2
22
33> Extend Claude Desktop on 3P with connectors, plugin marketplaces, organization plugins, skills, and hooks for administrators and end users
44
5Claude Desktop on third-party (3P) supports the same extensibility model as standard Claude Desktop ([MCP connectors](/docs/connectors/getting-started.md), [skills](/docs/skills/overview.md), and [plugins](/docs/plugins/overview.md)), with the key difference that administrators provision them through managed configuration and the filesystem rather than the claude.ai admin console.
5Claude Desktop on third-party (3P) supports the same extensibility model as standard Claude Desktop ([MCP connectors](/docs/connectors/getting-started.md), [skills](/docs/skills/overview.md), and [plugins](/docs/plugins/overview.md)), with the key difference that administrators provision them through managed configuration and the filesystem.
66
77There are three layers, in order of precedence:
88
from line 98
9898| Key | Default | What it does |
9999| - | - | - |
100100| `headersHelper` | None | Executable that prints the request headers as a flat JSON object to stdout. |
101| `headersHelperArgs` | None | List of arguments passed to the helper, one argument per item, in order. Arguments are visible in the app log and to other processes on the device, so don't put secrets in them. An app version that doesn't read this key runs the helper with no arguments. |
101102| `headersHelperTtlSec` | 300 | Seconds the returned headers stay valid. |
102103| `headersHelperRefreshBufferSec` | 60 | Seconds before expiry that the helper re-runs. Set it above the helper's typical runtime. |
103104
104The helper follows the [`inferenceCredentialHelper`](/docs/third-party/claude-desktop/credential-helper) execution model, with four differences: no arguments, a 30-second time limit, no `CLAUDE_HELPER_CONTEXT`, and no prompting for input. The helper applies only to servers provisioned through managed configuration and never replaces the `Authorization` header on `oauth` entries.
105The helper follows the [`inferenceCredentialHelper`](/docs/third-party/claude-desktop/credential-helper) execution model, with these differences: a 30-second time limit, no `CLAUDE_HELPER_CONTEXT`, and no prompting for input. The helper applies only to servers provisioned through managed configuration and never replaces the `Authorization` header on `oauth` entries.
105106
106107While the connection is open, the TTL schedule triggers renewal, and a request that the server rejects with HTTP 401 or 403 also re-runs the helper and, when it returns new headers, is retried once with them (Claude Desktop 1.46388.1 or later). A failed helper run does not interrupt the connection; Claude Desktop keeps the current headers and retries on its schedule. A failure while the server is connecting shows the server as needing authentication.
107108
from line 196
195196
196197### Configure the marketplace
197198
198You can add marketplaces directly in the [in-app configuration window](/docs/third-party/claude-desktop/in-app-configuration.md): in the **Plugins** section, click **Add marketplace** and choose **Blank**, **GitHub repo**, **Git URL**, or **Marketplace URL**. The form validates the entry and exports the encoded JSON for you.
199If your organization is managed from the Enterprise Admin Console, add marketplaces on the [**Plugins** page under **Desktop 3P**](/docs/third-party/claude-desktop/admin-console.md#plugin-marketplaces) in the console. Otherwise, you can add marketplaces directly in the [in-app configuration window](/docs/third-party/claude-desktop/in-app-configuration.md): in the **Plugins** section, click **Add marketplace** and choose **Blank**, **GitHub repo**, **Git URL**, or **Marketplace URL**. The form validates the entry and exports the encoded JSON for you.
199200
200201
201202
202203The Plugins section of the in-app configuration window, with the Add marketplace menu and the organization plugins folder.
203204
204To write the configuration yourself instead, add the repository to the [`allowedPluginMarketplaces`](/docs/third-party/claude-desktop/configuration#allowedpluginmarketplaces) configuration key. The key is read from an MDM profile, local configuration, or the [bootstrap server](/docs/third-party/claude-desktop/bootstrap.md) response. In an MDM profile the value is a JSON array encoded as a string (see [Value types](/docs/third-party/claude-desktop/configuration-format.md#value-types)). In a local configuration file or the bootstrap response the value is a native JSON array.
205To write the configuration yourself instead, add the repository to the [`allowedPluginMarketplaces`](/docs/third-party/claude-desktop/configuration#allowedpluginmarketplaces) configuration key. In an MDM profile the value is a JSON array encoded as a string (see [Value types](/docs/third-party/claude-desktop/configuration-format.md#value-types)). In a local configuration file or a [bootstrap server](/docs/third-party/claude-desktop/bootstrap.md) response, the value is a native JSON array.
205206
206207```xml .mobileconfig (macOS) theme={null}
207208<key>allowedPluginMarketplaces</key>
from line 334
333334
334335> [!NOTE]
335336>
336> Each entry in `org-plugins/` must carry a valid manifest: a `.claude-plugin/plugin.json`, or a top-level `SKILL.md` whose frontmatter declares `agents` or `mcpServers` (a skill folder that also acts as a plugin). A plain skill folder does not qualify on its own. To distribute a single skill, place it under `skills/<name>/SKILL.md` in a plugin that has a `plugin.json`. A directory with no valid manifest is not loaded and never appears in the user's plugin browser. The diagnostic report's plugin section shows the rejected entry and why. To distribute an MCP connector, declare it in a plugin's `.mcp.json` or use [`managedMcpServers`](#managed-mcp-servers-admin).
337> Each entry in `org-plugins/` must carry a valid manifest: a `.claude-plugin/plugin.json`, or a top-level `SKILL.md` whose frontmatter declares `agents` or `mcpServers` (a skill folder that also acts as a plugin). A plain skill folder does not qualify on its own. To distribute a single skill, place it under `skills/<name>/SKILL.md` in a plugin that has a `plugin.json`. A directory with no valid manifest is not loaded and never appears in the user's plugin browser. The diagnostic report's plugin section shows the rejected entry and why. To distribute an MCP connector, declare it in a plugin's `.mcp.json` or use [`managedMcpServers`](#managed-mcp-servers-admin). Claude Desktop doesn't start a desktop extension (`.mcpb`) that a plugin in `org-plugins/` names in its manifest.
337338
338339See the [plugins reference](https://code.claude.com/docs/en/plugins) for the full file format of each component, including the hooks schema.
339340
from line 395
394395
395396On a Windows device without Git for Windows, Claude Code runs a [hook](https://code.claude.com/docs/en/hooks) in PowerShell when the hook sets neither `shell` nor `args`. Hooks in Cowork sessions run the same way, on the device itself and not in Cowork's virtual machine. PowerShell rejects some commands that Bash accepts, such as a quoted program path followed by arguments.
396397
397A [`PreToolUse`](https://code.claude.com/docs/en/hooks#pretooluse) hook that runs this way and has no `onFailure` field blocks the tool call when it fails or times out. The hook error that Claude receives includes `blocking because Git Bash wasn't found`.
398A [`PreToolUse`](https://code.claude.com/docs/en/hooks#pretooluse) or [`PermissionRequest`](https://code.claude.com/docs/en/hooks#permissionrequest) hook that runs this way and has no `onFailure` field blocks the tool call when it fails or times out. The hook error that Claude receives includes `blocking because Git Bash wasn't found`.
398399
399400If your Windows devices have hooks, in a plugin or in Claude Code [managed settings](/docs/third-party/claude-desktop/code.md#interaction-with-claude-code%E2%80%99s-own-managed-settings), make one of these changes:
400401
401402* **Keep a hook written for Bash**: install Git for Windows in its default location (`C:\Program Files\Git`) on each device, then quit and reopen Claude Desktop
402* **Run the hook without a shell**: write the hook in [exec form](https://code.claude.com/docs/en/hooks#exec-form-and-shell-form), with the program in `command` and its arguments in `args`
403* **Run the hook without a shell**: write the hook in [exec form](https://code.claude.com/docs/en/hooks#exec-form-and-shell-form), with an executable such as an `.exe` (not a `.cmd` or `.bat` file) in `command` and its arguments in `args`
403404
404405After either change, a hook that fails or times out lets the tool call go ahead. To stop the tool call instead, add [`"onFailure": "block"`](https://code.claude.com/docs/en/hooks#block-the-action-when-a-hook-fails) to the hook.
405406
from line 424
423424
424425| Key | Default | Effect when `false` |
425426| - | - | - |
426| `isLocalDevMcpEnabled` | `true` | Users cannot add their own local MCP servers from **Settings → Developer**. |
427| `isLocalDevMcpEnabled` | `true` | Users cannot add their own local MCP servers from **Settings → Developer**. Unless [Claude Code's own managed settings](/docs/third-party/claude-desktop/code.md#interaction-with-claude-code%E2%80%99s-own-managed-settings) take precedence over the app's restrictions, local MCP servers bundled in marketplace plugins or in plugins users add themselves don't start. |
427428| `isDesktopExtensionEnabled` | `false` | Desktop extensions (`.mcpb`) bundled in plugins are not loaded. Set to `true` to allow them. |
428429| `isDesktopExtensionSignatureRequired` | `false` | (When `true`) Unsigned `.mcpb` extensions are rejected. |
429430| `skillCreationEnabled` | `true` | Users cannot create or upload skills in the app. Claude does not offer to create or update skills in conversations. |
from line 431
430431| `userPluginMarketplacesEnabled` | `true` | Users cannot add plugin marketplaces of their own, and the add-marketplace options are hidden. Marketplaces your organization did not [provision](#plugin-marketplaces-admin) are hidden too, and the app neither installs nor loads their plugins. The organization plugins directory, plugins users uploaded, and marketplaces you provision with `allowedPluginMarketplaces` are unaffected. |
431432| `userPluginUploadsEnabled` | `true` | Users cannot upload plugin files or create plugins with Claude; every in-app option for adding a plugin of their own is hidden. Plugins from your marketplaces and the organization plugins directory are unaffected. |
432433
433Setting `isLocalDevMcpEnabled` to `false` and leaving `isDesktopExtensionEnabled` at `false` restricts MCP servers and connectors to those delivered through `managedMcpServers` and `org-plugins/`, plus any that installed plugins bundle, whether from your marketplaces or added by users. To limit plugin-bundled servers to ones you name, or to none, set [`allowedPluginMcpServers`](/docs/third-party/claude-desktop/configuration#allowedpluginmcpservers) to a list of URL patterns. An empty list admits no plugin-bundled server. Setting [`skillCreationEnabled`](/docs/third-party/claude-desktop/configuration#skillcreationenabled) to `false` turns off skill creation and upload in the app. Skills already on the device keep working, as do skills from [organization plugins](#organization-plugins-admin). Users can still install plugins from the marketplaces you provision regardless of these settings. Setting `userPluginUploadsEnabled` to `false` removes only users' options for adding plugins themselves. Plugins a user uploaded earlier stay in place. See the [Locked down profile](/docs/third-party/claude-desktop/configuration.md#recommended-security-profiles) for a complete example.
434Setting `isLocalDevMcpEnabled` to `false` and leaving `isDesktopExtensionEnabled` at `false` restricts MCP servers and connectors to those delivered through `managedMcpServers` and `org-plugins/`, plus the remote servers that installed plugins bundle, whether from your marketplaces or added by users. In Code sessions, remote servers that users define in Claude Code's own configuration can also connect; see [Applied as managed policy](/docs/third-party/claude-desktop/code.md#applied-as-managed-policy) for the keys that stop them. On macOS and Windows, to keep a plugin's local server running while `isLocalDevMcpEnabled` is `false`, deliver the plugin through the [organization plugins directory](#organization-plugins-admin). To limit plugin-bundled servers to ones you name, or to none, set [`allowedPluginMcpServers`](/docs/third-party/claude-desktop/configuration#allowedpluginmcpservers) to a list of URL patterns. An empty list admits no plugin-bundled server. Setting [`skillCreationEnabled`](/docs/third-party/claude-desktop/configuration#skillcreationenabled) to `false` turns off skill creation and upload in the app. Skills already on the device keep working, as do skills from [organization plugins](#organization-plugins-admin). Users can still install plugins from the marketplaces you provision regardless of these settings. Setting `userPluginUploadsEnabled` to `false` removes only users' options for adding plugins themselves. Plugins a user uploaded earlier stay in place. See the [Locked down profile](/docs/third-party/claude-desktop/configuration.md#recommended-security-profiles) for a complete example.
434435
435436Setting [`userPluginMarketplacesEnabled`](/docs/third-party/claude-desktop/configuration#userpluginmarketplacesenabled) to `false` also hides every marketplace already on the device that your organization didn't [provision](#plugin-marketplaces-admin), whether a user added it in the app earlier or Claude Code registered it from a terminal on the same device. By default, Claude Code adds [Anthropic's official marketplace](https://code.claude.com/docs/en/plugins/anthropic-marketplaces) the first time a user starts an interactive terminal session, so that marketplace is hidden too unless you provision it. The app doesn't install or update plugins from a hidden marketplace. Plugins already installed from a hidden marketplace stay on the device, but they don't appear in the app or load in its sessions until you remove the key or set it back to `true`.
436437
third-party/claude-desktop/gateway Changed · +4 / -4 lines
from line 450
450450
451451### Single sign-on configuration keys
452452
453Single sign-on is enabled by setting `inferenceCredentialKind` to `interactive` **and** supplying `inferenceGatewayOidc`. Both are required — `interactive` alone (without `inferenceGatewayOidc`) selects a different mode where the gateway itself acts as the authorization server.
453Single sign-on is enabled by setting `inferenceCredentialKind` to `interactive` **and** supplying `inferenceGatewayOidc`. `interactive` alone (without `inferenceGatewayOidc`) selects a different mode where the gateway itself acts as the authorization server.
454454
455From Claude Desktop 2.7032.0, `inferenceCredentialKind: "external-idp"` with [`inferenceIdpOidc`](/docs/third-party/claude-desktop/configuration#inferenceidpoidc) and `inferenceIdpAuthFlow`, the keys [Amazon Bedrock identity provider sign-in](/docs/third-party/claude-desktop/bedrock.md#sign-in-with-your-identity-provider) uses, is an equivalent spelling of this configuration; `interactive` with `inferenceGatewayOidc` keeps working, so keep it until every device runs 2.7032.0 or later.
456
455457| Setting | MDM key | Required | Description |
456458| - | - | - | - |
457| Credential kind | `inferenceCredentialKind` | Yes — must be `interactive` | Selects sign-in instead of an API key. |
459| Credential kind | `inferenceCredentialKind` | Yes | `interactive` selects sign-in instead of an API key. |
458460| Gateway SSO IdP (OIDC) | `inferenceGatewayOidc` | Yes | A **single JSON object** describing the identity provider (fields below). The resulting token is sent to the gateway as the bearer credential. |
459461| Sign-in flow | `inferenceGatewayOidcAuthFlow` | No | `browser` (the default) runs the sign-in in the system browser. `broker` runs it through the [OS identity broker](/docs/third-party/claude-desktop/entra-broker.md) on Windows and macOS, which requires `issuer` to be a Microsoft Entra ID issuer (`https://login.microsoftonline.com/TENANT_ID/v2.0`) and needs no loopback redirect. |
460462
from line 490
488490<key>inferenceGatewayOidc</key>
489491<string>{"issuer":"https://YOUR_ORG.okta.com","clientId":"YOUR_CLIENT_ID","redirectPort":53180}</string>
490492```
491
492From Claude Desktop 2.7032.0, `inferenceCredentialKind: "external-idp"` with [`inferenceIdpOidc`](/docs/third-party/claude-desktop/configuration#inferenceidpoidc) and `inferenceIdpAuthFlow`, the keys [Amazon Bedrock identity provider sign-in](/docs/third-party/claude-desktop/bedrock.md#sign-in-with-your-identity-provider) uses, is an equivalent spelling of this configuration; `interactive` with `inferenceGatewayOidc` keeps working, so keep it until every device runs 2.7032.0 or later.
493493
494494Earlier app versions used `inferenceGatewayAuthScheme: "sso"` to select this mode. That value is deprecated, so set `inferenceCredentialKind: "interactive"` instead. The [configuration changelog](/docs/third-party/claude-desktop/configuration-changelog.md#v2-26454-0) says which app versions read the deprecated value.
495495
third-party/claude-desktop/installation Changed · +4 / -4 lines
from line 94
9494
9595## Verifying the deployment
9696
97On any configured device, open Claude Desktop, go to **Help → Troubleshooting → Generate Diagnostic Report**, and click **Export to file**. In the saved `.zip` file, `managed-config.txt` shows where the configuration was read from and every key the app applied, with secret values redacted and anything it could not parse listed under `Parse errors`. `provider-status.txt` shows whether the provider settings are complete and valid, and `deployment-mode.txt` shows whether the app is running in third-party mode.
97On any configured device, open Claude Desktop, go to **Help → Troubleshooting → Generate Diagnostic Report**, and click **Export to file**. In the saved `.zip` file, `managed-config.txt` shows where the configuration was read from and each key's configured or default value, with secret values redacted and anything it could not parse listed under `Parse errors`. `provider-status.txt` shows whether the provider settings are complete and valid, and `deployment-mode.txt` shows whether the app is running in third-party mode.
9898
99Also confirm that the [in-app configuration window](/docs/third-party/claude-desktop/in-app-configuration.md) (**Developer → Configure Third-Party Inference…**) opens read-only on a managed device. The app reads managed keys from the profile by name and silently ignores a misspelled key rather than reporting an error. On macOS, a window that is still editable means no recognized key reached the app, even if your MDM shows the profile as delivered. On Windows, even a misspelled value under `HKLM\SOFTWARE\Policies\Claude` counts as machine policy and locks the window, so check `managed-config.txt` in the diagnostic report to see which keys were actually read. If your profile deliberately sets [only app-behavior keys](/docs/third-party/claude-desktop/mdm.md#update-keys-and-managed-precedence) (the update, relaunch window, configuration re-check, or network proxy keys), an editable window is expected.
99Also confirm that the [in-app configuration window](/docs/third-party/claude-desktop/in-app-configuration.md) (**Developer → Configure Third-Party Inference…**) opens read-only on a managed device. The app reads managed keys from the profile by name and silently ignores a misspelled key rather than reporting an error. On macOS, a window that is still editable means no recognized key reached the app, even if your MDM shows the profile as delivered. On Windows, even a misspelled value under `HKLM\SOFTWARE\Policies\Claude` counts as machine policy and locks the window, so compare each key's value in `managed-config.txt` with the value you set. If your profile deliberately sets [only app-behavior keys](/docs/third-party/claude-desktop/mdm.md#update-keys-and-managed-precedence) (the update, relaunch window, configuration re-check, or network proxy keys), an editable window is expected.
100100
101101If the app shows the standard claude.ai sign-in screen instead of Cowork, the configuration was not read. Common causes:
102102
from line 103
103103* `inferenceProvider` is missing or its name is misspelled, or its value in the local configuration file or in Windows policy under `HKCU` isn't one the app recognizes
104104* The configuration was applied while the app was running (fully quit and relaunch)
105105* The configuration was written to the local config file but you're checking the managed location (or vice versa)
106* On Windows (v1.19367.0 and later), the configuration is in `HKCU\SOFTWARE\Policies\Claude` but machine policy is also present: any `REG_SZ`, `REG_EXPAND_SZ`, or `REG_DWORD` value directly under `HKLM\SOFTWARE\Policies\Claude` causes the app to ignore user policy entirely. In the diagnostic report (**Help → Troubleshooting → Generate Diagnostic Report**, then **Export to file**), `managed-config.txt` shows which keys the app read and lists under `Parse errors` any machine-policy values it could not use. A `REG_EXPAND_SZ` value shows as present in `reg query` output while the app reports the managed configuration as invalid or absent, because the app counts the value as machine policy but cannot read its contents
106* On Windows (v1.19367.0 and later), the configuration is in `HKCU\SOFTWARE\Policies\Claude` but machine policy is also present: any `REG_SZ`, `REG_EXPAND_SZ`, or `REG_DWORD` value directly under `HKLM\SOFTWARE\Policies\Claude` causes the app to ignore user policy entirely. In the diagnostic report (**Help → Troubleshooting → Generate Diagnostic Report**, then **Export to file**), `managed-config.txt` shows each key's configured or default value and lists under `Parse errors` any machine-policy values it could not use. A `REG_EXPAND_SZ` value shows as present in `reg query` output while the app reports the managed configuration as invalid or absent, because the app counts the value as machine policy but cannot read its contents
107107
108108If the app shows a configuration error instead, it read the configuration but can't use it. For example, a required key for the chosen provider is missing or invalid, or a macOS configuration profile, Windows policy under `HKLM`, or the root-owned policy file on Linux sets [`inferenceProvider`](/docs/third-party/claude-desktop/configuration#inferenceprovider) to a value the app doesn't recognize. Check **Help → Troubleshooting** or the application log at `~/Library/Logs/Claude-3p/main.log` (macOS) / `%LOCALAPPDATA%\Claude-3p\Logs\main.log` (Windows).
109109
third-party/claude-desktop/bootstrap Changed · +2 / -1 lines
from line 47
4747
4848Run the endpoint across multiple replicas or regions behind a load balancer. Do not rely on response caching for availability: responses are per-user and carry credentials (see the `Cache-Control: no-store` guidance under [Server responsibilities](#server-responsibilities)). If your configuration data lives in a database, a read replica of that store improves availability without caching responses.
4949
50When a refetch returns different values, a few settings apply to the running app without a restart, such as the banner, the display name, and the token limit, as do the two lifecycle keys described below. Every other setting, including inference credentials, the egress allowlist, MCP servers, and the model list, stays on the values the app launched with. For those, the app prompts the user to restart and applies the new response when it relaunches.
50When two refetches in a row return the same changed values, a few settings apply to the running app without a restart, such as the banner, the display name, and the token limit, as do the two lifecycle keys described below. Every other setting, including inference credentials, the egress allowlist, MCP servers, and the model list, stays on the values the app launched with. For those, the app prompts the user to restart and applies the new response when it relaunches.
5151
5252Claude Desktop 1.40609.0 and later enforce that restart. Once a background re-check returns a changed response that needs a restart, the user can keep working for [`relaunchEnforcementHours`](/docs/third-party/claude-desktop/configuration#relaunchenforcementhours) (24 hours by default, at most 336 hours, or `0` to require the restart at once; releases before 1.46388.1 default to 1 hour). After that window the app blocks further use until it restarts, and it relaunches on its own once it has been idle for two minutes (no Claude task running and no keyboard or pointer input).
5353
from line 514
514514| Silent refresh fails after \~1 hour with `AADSTS90009` | `scopes` uses the `api://CLIENT_ID/.default` form. Use the bare-GUID `CLIENT_ID/.default` form. |
515515| Some keys you returned are not applied | They failed schema validation, are structurally excluded, or were dropped by origin pinning. If the app instead shows an **Apply settings from your organization?** dialog, the whole response is waiting for [user consent](#keys-that-require-user-consent) and none of it has been applied yet. The desktop log (`~/Library/Logs/Claude-3p/main.log` on macOS, `%LOCALAPPDATA%\Claude-3p\logs\main.log` on Windows) records which keys were dropped and why. |
516516| Browser opens to your identity provider's device page instead of yours | In device-code mode, `verification_uri` must share the `bootstrapUrl` origin. Federate behind your own page. |
517| Changes on your server don't reach running apps, which show no restart prompt | A running app acts on a changed response once two [background re-checks](#how-it-works) in a row return the same values. A response that differs only in `expiresAt` counts as unchanged. If your server returns a different response to every request, for example a new credential each time, a running app applies none of the changes until it relaunches, and `main.log` records `served configuration changed again before it was confirmed`. |
517518
third-party/claude-desktop/browser Changed · +1 / -1 lines
from line 22
2222
2323## Built-in browser
2424
25When the built-in browser is on, Cowork and Code sessions show a **Browser** pane where users can open and sign in to any site. In the pane, Claude can open pages, read them, click, type, fill in forms, take screenshots, and run JavaScript on a page.
25When the built-in browser is on, Cowork and Code sessions show a **Browser** pane where users can open and sign in to any site. Except in an [SSH remote session](/docs/third-party/claude-desktop/ssh-remote-sessions.md), Claude can use the pane to open pages, read them, click, type, fill in forms, take screenshots, and run JavaScript on a page.
2626
2727Before Claude acts on a site, Claude Desktop asks the user to approve it. **Allow once** covers that page, and **Always allow** saves the site on the device until the user revokes it in **Settings**. After a user approves a site, Claude can read and interact with its pages as the user sees them, including pages behind the user's sign-in on that site.
2828
third-party/claude-desktop/connectors-github Changed · +1 / -1 lines
from line 15
1515| Credential handling | Token delivered through the entry's `headers` or a headers helper script | User signs in; the token is acquired and stored encrypted on the device |
1616| GitHub Enterprise Server | Not available; github.com only | Supported; set `host` |
1717| Tool surface controls | Per-tool [`toolPolicy`](/docs/third-party/claude-desktop/configuration#managedmcpservers) | `toolsets`, `readOnly`, and per-tool [`toolPolicy`](/docs/third-party/claude-desktop/configuration#managedmcpservers) |
18| Claude Desktop version | Any version that supports managed MCP servers | Requires a version that includes the bundled server (beta) |
18| Claude Desktop version | Any version that supports managed MCP servers | Requires a version that includes the bundled server |
1919
2020## Remote connector
2121
third-party/claude-desktop/data-storage Changed · +1 / -1 lines
from line 30
3030| - | - |
3131| `ant-did` | The random device identifier described above. |
3232| `configLibrary/` | Locally authored configuration (from the in-app configuration window). `_meta.json` records which saved configuration is applied; each is a `<id>.json` file alongside it. Ignored when a managed profile is present. |
33| `local-agent-mode-sessions/.../cowork_account_settings.json` | User-level preferences set in the app (display name, locale, memory toggle). |
33| `local-agent-mode-sessions/.../cowork_account_settings.json` | User-level preferences set in the app, such as locale, the memory toggle, and the text entered under **Instructions for Claude** and **Global instructions**. |
3434| `local-agent-mode-sessions/` | Cowork and Chat conversation history. One `local_<uuid>.json` file plus a working directory per session, scoped by the `ant-did` device identifier and the organization ID, not by who is signed in to the app. The working directory includes an `uploads/` subdirectory with copies of files attached to the conversation and an `outputs/` subdirectory for files Claude creates. |
3535| `local-agent-mode-sessions/.../memory/` | Cowork memory: a `CLAUDE.md` instructions file plus a `memory/` subdirectory of Markdown notes Claude writes about the user's preferences, projects, and feedback. See [Memory](#memory). |
3636| `local-agent-mode-sessions/.../spaces/<projectId>/memory/` | Markdown memory notes for one project, used by Cowork sessions and Chat conversations inside that project. See [Memory](#memory). |
third-party/claude-desktop/foundry Changed · +2 / -2 lines
from line 72
7272
7373If the app can no longer renew the credential silently, it shows a **Sign in again** prompt; clicking it reopens the configured sign-in flow. For the device-code and browser flows this happens when the stored refresh token expires or is revoked. For the broker flow it happens when the broker can no longer renew the token silently.
7474
75`inferenceFoundryTenantId`, `inferenceFoundryClientId`, and `inferenceFoundryAuthFlow` can be set through an MDM profile or a [bootstrap server](/docs/third-party/claude-desktop/bootstrap.md). When a bootstrap server delivers `inferenceFoundryResource`, `inferenceFoundryTenantId`, or `inferenceFoundryClientId`, the values are among the [keys that require user consent](/docs/third-party/claude-desktop/bootstrap.md#keys-that-require-user-consent), so users may see a one-time approval dialog depending on how `bootstrapUrl` reached the device.
75`inferenceFoundryTenantId`, `inferenceFoundryClientId`, and `inferenceFoundryAuthFlow` can be set through the [Enterprise Admin Console](/docs/third-party/claude-desktop/admin-console.md#choose-how-users-authenticate-to-your-provider), an MDM profile, or a [bootstrap server](/docs/third-party/claude-desktop/bootstrap.md). When a bootstrap server delivers `inferenceFoundryResource`, `inferenceFoundryTenantId`, or `inferenceFoundryClientId`, the values are among the [keys that require user consent](/docs/third-party/claude-desktop/bootstrap.md#keys-that-require-user-consent), so users may see a one-time approval dialog depending on how `bootstrapUrl` reached the device.
7676
7777> [!NOTE]
7878>
from line 220
220220
221221For broker-flow sign-in failures (error codes `AADSTS50011`, `AADSTS900971`, `AADSTS7000218`, or a message that the OS identity broker is unavailable), see [Troubleshoot](/docs/third-party/claude-desktop/entra-broker.md#troubleshoot) on the OS identity broker page. To unblock a device that cannot meet the broker requirements, set `inferenceFoundryAuthFlow` to `browser` for that device instead.
222222
223Each sign-in attempt has a time limit: five minutes for the device-code and broker flows and two minutes for the browser flow. If the user does not finish within the limit, the attempt fails and the user can click **Sign in with Microsoft** to start again.
223Each sign-in attempt has a time limit: five minutes for the device-code and broker flows and ten minutes for the browser flow. If the user does not finish within the limit, the attempt fails and the user can click **Sign in with Microsoft** to start again.
224224
third-party/claude-desktop/import Changed · +1 / -1 lines
from line 13
1313Along with the sessions, the automatic copy brings over other things you had under the earlier configuration:
1414
1515* **Projects**: your Cowork projects. New tasks use a project's instructions only after you accept them.
16* **Memory and instructions**: your Cowork memory and global instructions. New tasks use them right away. Your **Instructions for Claude** (**Settings → General**) come over too. Instructions you already set under the new configuration are kept.
16* **Memory and instructions**: your Cowork memory, unless you paused it, and your global instructions. New tasks use them right away. Your **Instructions for Claude** (**Settings → General**) come over too. Instructions you already set under the new configuration are kept.
1717* **Scheduled tasks**: your scheduled tasks from Cowork and Code. Each stays on or off as it was.
1818* **Uploaded plugins**: plugins you uploaded, unless your administrator has turned off plugin uploads ([`userPluginUploadsEnabled`](/docs/third-party/claude-desktop/configuration#userpluginuploadsenabled)). Each stays on or off as it was.
1919* **Saved skills**: skills you saved, unless your administrator has turned off skill creation ([`skillCreationEnabled`](/docs/third-party/claude-desktop/configuration#skillcreationenabled)).
third-party/claude-desktop/mdm Changed · +1 / -1 lines
from line 141
141141
142142When a managed source sets only keys from this group (any of them), the device keeps its locally authored configuration and the configuration window stays editable. The group is still enforced as a unit: every key in it is resolved from the managed source alone, so a locally set or bootstrap-served value for any of them is ignored even if the profile sets only one. A profile that sets the update keys but not `relaunchEnforcementHours` or `configRecheckIntervalMinutes` therefore runs their defaults (24 hours and 10 minutes) on those devices, so set them in the same profile when you want other values.
143143
144If the managed profile sets any other recognized key, the normal rule above applies and the whole configuration is managed.
144If the managed profile sets any other recognized key, the normal rule above applies and the whole configuration is managed. For a profile that also sets `bootstrapUrl`, see [Which configuration source applies on a device](/docs/third-party/claude-desktop/configuration-format.md#which-configuration-source-applies-on-a-device).
145145
146146## 5. Distribute the app
147147
third-party/claude-desktop/network-proxy Changed · +2 / -2 lines
from line 62
6262}
6363```
6464
65In a macOS configuration profile the same key is a `<key>egressProxyUrl</key><string>http://proxy.example.com:8080</string>` pair in the `com.anthropic.claudefordesktop` payload, and on Windows it is a `REG_SZ` value named `egressProxyUrl` under `HKLM\SOFTWARE\Policies\Claude`. See [Value types](/docs/third-party/claude-desktop/configuration-format.md#value-types) and [Deploy with MDM](/docs/third-party/claude-desktop/mdm.md) for the surrounding profile and registry structure.
65In a macOS configuration profile the same key is a `<key>egressProxyUrl</key><string>http://proxy.example.com:8080</string>` pair in the `com.anthropic.claudefordesktop` payload. On Windows it is a `REG_SZ` value named `egressProxyUrl` under `HKLM\SOFTWARE\Policies\Claude`, or under `HKCU\SOFTWARE\Policies\Claude` if the rest of your registry policy is there. The app [ignores `HKCU` policy when machine policy is present](/docs/third-party/claude-desktop/mdm#4-deploy-the-configuration). See [Value types](/docs/third-party/claude-desktop/configuration-format.md#value-types) and [Deploy with MDM](/docs/third-party/claude-desktop/mdm.md) for the surrounding profile and registry structure.
6666
6767Three behaviors to plan for:
6868
from line 172
172172
173173## Troubleshoot
174174
175**"Can't reach" banner.** The app could not get a response from the inference host the banner names. Behind a proxy this usually means the proxy resolved for that host is unreachable, requires a sign-in the app cannot perform, or does not allow the host. Use **Copy report** on the banner to capture the details, confirm the device's OS proxy settings (or the pinned key) point at a reachable HTTP proxy, and confirm the proxy allows the inference host listed in [Required egress paths](/docs/third-party/claude-desktop/telemetry.md#required-egress-paths).
175**"Can't reach" banner.** The app could not get a response from the inference host the banner names. Behind a proxy this usually means the proxy resolved for that host is unreachable, requires a sign-in the app cannot perform, or does not allow the host. Use **Copy report for IT** on the banner to capture the details, confirm the device's OS proxy settings (or the pinned key) point at a reachable HTTP proxy, and confirm the proxy allows the inference host listed in [Required egress paths](/docs/third-party/claude-desktop/telemetry.md#required-egress-paths).
176176
177177**Confirm which proxy is in effect.** Open `main.log` in the app's [logs directory](/docs/third-party/claude-desktop/data-storage.md). When a key is pinned, startup logs `[egress-proxy] pinned to fixed proxy at proxy.example.com:8080; OS proxy settings ignored` (or `pinned to PAC script at …`). The log records the proxy handed to the agent the first time and whenever it changes, for example `Resolved system proxy for Code sessions: http://proxy.example.com:8080`, a `Keeping system proxy` line when the app reuses the last proxy because the OS answer was direct, SOCKS, or missing, and a `Skipping SOCKS proxy entry` line if the OS answer was SOCKS. If Claude Code managed settings supplied any proxy variable, a line names the variables it set or replaced and the settings file they came from. For Cowork's sandboxed shell, `cowork_vm_node.log` in the same directory records `[VM:start] guest egress pinned to fixed proxy at …` (or `PAC script at …`) when the sandbox starts with a key pinned, and `[VM:start] PAC fetch from … failed (…); guest connects directly` if the sandbox's copy of the script could not be downloaded.
178178
third-party/claude-desktop/overview Changed · +1 / -1 lines
from line 83
8383
8484> **[Installation and setup](/docs/third-party/claude-desktop/installation.md)**
8585>
86> Roll out Claude Desktop on 3P to your organization with MDM, or configure a single machine for evaluation.
86> Roll out Claude Desktop on 3P to your organization, or configure a single machine for evaluation.
8787
8888> **[Configuration reference](/docs/third-party/claude-desktop/configuration.md)**
8989>
third-party/claude-desktop/telemetry Changed · +1 / -1 lines
from line 84
8484
8585Every record sent to your collector carries the user's identity as two resource attributes, on all three `service.name` streams:
8686
87* `enduser.id` — the signed-in user's identity. With an interactive sign-in flow (for example, Workforce Identity Federation or Google sign-in on Google Cloud's Agent Platform), this is the identity from the provider's claims, normally the user's email address. With credential methods that carry no identity claims (a static key, a credential helper, or an application default credentials file), it is the operating-system login name.
87* `enduser.id` — the signed-in user's identity. With a sign-in that carries identity claims (for example, Workforce Identity Federation or Google sign-in on Google Cloud's Agent Platform), this is the identity from the provider's claims, normally the user's email address. Some credential methods carry no identity claims, such as a static key, a credential helper, an application default credentials file, [in-app AWS sign-in](/docs/third-party/claude-desktop/bedrock.md#in-app-aws-sign-in), or a [named AWS profile](/docs/third-party/claude-desktop/bedrock.md#named-profile). With one of those, on a device managed from the [Enterprise Admin Console](/docs/third-party/claude-desktop/admin-console.md), it is the email address of the user's Claude account. On other devices, it is the operating-system login name.
8888* `process.owner` — the operating-system login name.
8989
9090`enduser.id` is the same identity the app shows in the sidebar and account menu, and is controlled by the [`endUserAttribution`](/docs/third-party/claude-desktop/configuration#enduserattribution) key: set it to `false` to remove the identity from both the app and the export. `process.owner` is not gated by that key — it is standard OpenTelemetry process metadata and is always present. A static value set under [`otlpResourceAttributes`](/docs/third-party/claude-desktop/configuration#otlpresourceattributes) overrides either attribute: a static `enduser.id` is always passed through — taking precedence over the signed-in identity, and surviving `endUserAttribution: false` — and a static `process.owner` replaces the login name.
third-party/claude-desktop/web-tools Changed · +1 / -1 lines
from line 95
9595}
9696```
9797
98Set the per-entry `toolPolicy` to `"allow"` so users aren't prompted to approve each search. `headersHelper` is an executable that prints the auth header as a JSON object to stdout; it follows the same execution model as [`inferenceCredentialHelper`](/docs/third-party/claude-desktop/credential-helper) (exit 0, stdout read as JSON) but always runs with no arguments, and the output here is a flat header map, not the `{token, headers}` shape `inferenceCredentialHelper` uses.
98Set the per-entry `toolPolicy` to `"allow"` so users aren't prompted to approve each search. `headersHelper` is an executable that prints the auth header as a JSON object to stdout; it follows the same execution model as [`inferenceCredentialHelper`](/docs/third-party/claude-desktop/credential-helper) (exit 0, stdout read as JSON), but the output here is a flat header map, not the `{token, headers}` shape `inferenceCredentialHelper` uses.
9999
100100| Provider | Header your script should output |
101101| - | - |