HomeDocs changes
Documentation changes
What the docs quietly changed.
Anthropic's published documentation, fetched on a schedule and diffed against the last copy. Every edit dated, including the ones that never made a release note.
Last poll read all 4 cleanly: Claude Code CLI, Claude Developer Platform, Claude Documentation, Model Context Protocol. A date here is when this site noticed, not when Anthropic edited.
Claude Developer Platform
820pages tracked
773 changed this week · last read 23 Oct 2026
Claude Code CLI
228pages tracked
147 changed this week · last read 23 Oct 2026
Claude Documentation
280pages tracked
77 changed this week · last read 23 Oct 2026
Model Context Protocol
357pages tracked
171 changed this week · last read 23 Oct 2026
100 on this page
ReadSourcePageWhatLines
10 Sep 2026
634 pages
+345362 −331038 lines
claude-docs
claude-science/tools-and-environments
±changed
+1−1
claude-docschanged
claude-science/tools-and-environments
Claude writes and runs Python, R, and shell commands. Python and R run in a persistent kernel that keeps variables in memory across steps in a session. The kernel ends after about
Claude writes and runs Python, R, and shell commands (PowerShell on Windows). Python and R run in a persistent kernel that keeps variables in memory across steps in a session. The
+1 −1 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/run-on-windows-wsl
−removed
+0−90
claude-docsremoved
claude-science/run-on-windows-wsl
No line of the body moved in this read.
+0 −90 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/run-on-windows
+new page
+56−0
claude-docsnew page
claude-science/run-on-windows
No line of the body moved in this read.
+56 −0 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/run-on-remote-linux-server
±changed
+11−11
claude-docschanged
claude-science/run-on-remote-linux-server
Sign in with your Claude account. If the sign-in redirect can't find its way back through the tunnel, choose **Paste a code** on the sign-in screen. Then complete the setup wizard
Sign in with your Claude account. If the sign-in redirect can't find its way back through the tunnel, choose **Paste code instead** on the sign-in screen. Then complete the setup w
| Symptom | What it means
| ----------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------
| `command not found: claude-science` | `~/.local/bin` isn't on your PATH yet. Run `. ~/.profile` or open a new terminal
+11 −11 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/overview
±changed
+3−2
claude-docschanged
claude-science/overview
Claude Science is a desktop application that pairs Claude with an analysis environment on your computer. Available in beta on macOS and Linux.
Claude Science is a desktop application that pairs Claude with an analysis environment on your computer. Available in beta on macOS, Windows, and Linux.
* macOS 13 or later (Apple silicon or Intel), or Linux x64 on a glibc-based distribution.
* macOS 13 or later (Apple silicon or Intel), Windows 11 (x64), or Linux x64 on a glibc-based distribution.
* On Windows: the Microsoft Visual C++ Redistributable (x64), which some features depend on and most PCs already have (see [Run on Windows](/docs/claude-science/run-on-windows#requ
+3 −2 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/network-requirements
±changed
+10−9
claude-docschanged
claude-science/network-requirements
| Domain | Required when | Purpose
| --------------------------------------- | --------------------------------------------------- | ----------------------------------------------------------------------------------
| `claude.ai` | Always | Browser-based sign-in, usage analytics, feature configuration, and the catalog of
| `platform.claude.com` | Always | Completing sign-in (the OAuth token exchange)
| `api.anthropic.com` | Always | The Claude API for every request Claude makes, plus account and usage information
+10 −9 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/manage-on-devices
±changed
+7−5
claude-docschanged
claude-science/manage-on-devices
Configuration: \~/.claude-science/config.toml holds all app settings. Every key is optional; the app starts with no file present. This is the file to deploy through device manageme
Configuration: config.toml in the app's default data folder (`~/.claude-science/config.toml` on macOS and Linux, `%USERPROFILE%\.claude-science\config.toml` on Windows) holds all a
Authentication tokens and the shared package environment live under \~/.claude-science/ regardless of the data directory, so endpoint backup or wipe policies that target the data d
Authentication tokens and the shared package environment live under the default data folder (`~/.claude-science/`, or `%USERPROFILE%\.claude-science\` on Windows) regardless of the
+7 −5 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/get-started
±changed
+38−25
claude-docschanged
claude-science/get-started
> Install Claude Science on macOS or Linux, sign in with your Claude account, and run your first analysis.
> Install Claude Science on macOS, Windows, or Linux, sign in with your Claude account, and run your first analysis.
macOS: Download the installer from claude.com/product/claude-science and double-click to install. On first launch, the app sets up its runtime and starter Python and R environments
<Tabs>
<Tab title="macOS">
+38 −25 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/custom-connectors
±changed
+1−1
claude-docschanged
claude-science/custom-connectors
Local-command connectors run inside the sandbox with the same network limits as Claude's code and a per-connector writable directory. Environment variables for local connectors are
Local-command connectors run inside the sandbox with the same network limits as Claude's code and a per-connector writable directory. Environment variables for local connectors are
+1 −1 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/corporate-networks
±changed
+38−30
claude-docschanged
claude-science/corporate-networks
| Network shape | macOS
| ---------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------
| Explicit outbound HTTP proxy (HTTP CONNECT) | Supported; detected automatically from the system settings
| Proxy that requires Basic authentication | Supported, with credentials in the proxy address
| Proxy that requires NTLM, Negotiate, or Kerberos authentication | Not supported
+38 −30 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/configuration-file-reference
±changed
+16−13
claude-docschanged
claude-science/configuration-file-reference
Claude Science reads optional settings from a TOML file at `~/.claude-science/config.toml` on macOS and Linux. Every key has a default, so the app starts with no file present; admi
Claude Science reads optional settings from a TOML file named `config.toml` in its default data folder, which is `~/.claude-science/config.toml` on macOS and Linux and `%USERPROFIL
On Windows, save the file as UTF-8 and write Windows paths inside it in single quotes (for example `'C:\ProgramData\corp\corporate-ca.pem'`) or with forward slashes, because a back
| Key | Type | Default | Effect
+16 −13 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/compute-providers
±changed
+1−1
claude-docschanged
claude-science/compute-providers
In Settings > Compute, under Model endpoints, click Connect on NVIDIA BioNeMo NIM. Import the skills from the BioNeMo Agent Toolkit, add your NVIDIA NGC API credential, and connect
Model endpoints are available in Claude Science on macOS and Linux. In Settings > Compute, under Model endpoints, click Connect on NVIDIA BioNeMo NIM. Import the skills from the Bi
+1 −1 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/command-line-settings
±changed
+3−2
claude-docschanged
claude-science/command-line-settings
claude-science serve starts Claude Science and opens the web app in your browser at a single-use login link. Everyday use is that one command. The others manage the running program
claude-science serve starts Claude Science and opens the web app in your browser at a single-use login link. Everyday use is that one command. The others manage the running program
| `claude-science uninstall` | Windows only. Remove the app, its shortcuts, and its PATH entry while keeping your data; `--purge` also deletes the data directory. Quit Cla
These two work on every command.
These two work on every command. On Windows, `~` in the defaults below is your user profile folder, `%USERPROFILE%`.
+3 −2 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/changelog
±changed
+5−0
claude-docschanged
claude-science/changelog
<Update label="0.1.47" description="September 10, 2026">
* Claude Science is now available for Windows: download it from [claude.com/product/claude-science](https://claude.com/product/claude-science), and see [Run on Windows](/docs/cla
* Various bug fixes and improvements
</Update>
+5 −0 lines
Recorded here 10 Sep 2026 18:09 UTC
claude-docs
claude-science/run-on-remote-linux-server
±changed
+10−10
claude-docschanged
claude-science/run-on-remote-linux-server
| Symptom | What it means
| ----------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------
| `command not found: claude-science` | `~/.local/bin` isn't on your PATH yet. Run `. ~/.profile` or open a new terminal
| An error mentioning `bwrap too old` | The server's bubblewrap is older than 0.8.0. Upgrade it, or use a distribution t
| An error mentioning `cannot create unprivileged user namespaces` | The kernel or an AppArmor profile blocks the sandbox from creating user namespac
+10 −10 lines
Upstream edited 10 Sep 2026 18:09 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
claude-science/configuration-file-reference
±changed
+18−18
claude-docschanged
claude-science/configuration-file-reference
| Key | Type | Default | Effect
| ----------------- | ------------------------------------ | -------- | ------------------------------------------------------------------------------------------------------------
| `proxy` | string (URL) | unset | The outbound proxy for the app's HTTP and HTTPS connections, for example `http://proxy.example.corp:8080`. S
| `no_proxy` | string (comma-separated) | unset | Hosts that bypass the proxy, as exact hostnames or domain suffixes in one comma-separated string, for exampl
| `ca_bundle` | string (absolute path) | unset | PEM file added to the app's default TLS trust for sign-in, the Claude API, Anthropic-hosted connectors, upda
+18 −18 lines
Upstream edited 10 Sep 2026 18:09 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
claude-science/command-line-settings
±changed
+28−28
claude-docschanged
claude-science/command-line-settings
| Command | What it does
| ----------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------
| `claude-science serve` | Start the background program and open the web app in your browser at a single-use login link. One runs per data directory; Ctrl-C stops it.
| `claude-science open` | Mint a fresh login link from the running program and open it in your browser.
| `claude-science url` | Print a fresh login link alone on standard output and nothing else.
+28 −28 lines
Upstream edited 10 Sep 2026 18:09 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-code
workflows
±changed
+1−1
claude-codechanged
workflows
`/effort ultracode` lasts for the current session; to have every session start with it, set the [`ultracode`](/docs/en/settings-reference#ultracode) setting. Drop back with `/effor
`/effort ultracode` lasts for the current session; to have every session start with it, set the [`ultracode`](/docs/en/settings-reference#ultracode) setting. Drop back with `/effor
+1 −1 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
vs-code
±changed
+5−3
claude-codechanged
vs-code
* **Model**: select **Switch model…** from the command menu to change the model mid-session. You can also click the model name at the bottom of the prompt box to open the same pick
* **Model**: select **Switch model…** from the command menu to change the model mid-session. You can also click the model name at the bottom of the prompt box to open the same pick
You can also hold `Shift` while dragging files into the prompt box to add them as attachments. Click the X on any attachment to remove it from context.
To attach an image, paste it from your clipboard into the prompt box. You can also hold `Shift` while dragging files into the prompt box to add them as attachments. Click the X on
Click the **Session history** button at the top of the Claude Code panel to access your conversation history. You can search by keyword or browse by time. Click any conversation to
+5 −3 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
settings-reference
±changed
+6−3
claude-codechanged
settings-reference
Cap the [effort level](/docs/en/model-config#adjust-effort-level) a session can use, leaving lower levels available. Any higher level runs at the cap instead, including one from `/
Cap the [effort level](/docs/en/model-config#adjust-effort-level) a session can use, leaving lower levels available. Any higher level runs at the cap instead, including one from `/
* **Effect on ultracode**: a cap below `xhigh` makes [ultracode](#ultracode) unavailable on the models the cap applies to
* **Type**: array of objects, each with `marketplace` and `plugin` strings
* **Type**: array of objects, each with `marketplace` and `plugin` strings. An entry can instead be a `"plugin@marketplace"` string such as `"telegram@claude-plugins-official"`, wh
+6 −3 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
remote-control
±changed
+1−1
claude-codechanged
remote-control
All traffic travels through the Anthropic API over TLS, the same transport security as any Claude Code session. The connection uses multiple short-lived credentials, each scoped to
All traffic travels through the Anthropic API over TLS, the same transport security as any Claude Code session. The connection uses multiple short-lived credentials, each scoped to
+1 −1 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
model-config
±changed
+10−2
claude-codechanged
model-config
When ultracode isn't available, for example when [workflows are turned off](/docs/en/workflows#turn-workflows-off), `--effort ultracode` sets `xhigh` effort only.
<span id="when-ultracode-is-available" />
Ultracode is unavailable when:
+10 −2 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
managed-settings
±changed
+3−3
claude-codechanged
managed-settings
| `allowedHttpHookUrls` | Claude Code enforces an empty [allowlist](/docs/en/settings-reference#allowedhttphookurls) until you fix the value. If only an individual entry is
| `httpHookAllowedEnvVars` | Claude Code enforces an empty [allowlist](/docs/en/settings-reference#httphookallowedenvvars) until you fix the value. If only an individual entry
| `allowedHttpHookUrls` | Claude Code enforces an empty managed [allowlist](/docs/en/settings-reference#allowedhttphookurls) until you fix the value, so an HTTP hook runs o
| `httpHookAllowedEnvVars` | Claude Code enforces an empty managed [allowlist](/docs/en/settings-reference#httphookallowedenvvars) until you fix the value, so a header variabl
Before v2.1.267, Claude Code dropped `allowedHttpHookUrls`, `httpHookAllowedEnvVars`, or `allowedChannelPlugins` whole when the value or any one entry was invalid, and behaved as i
+3 −3 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
errors
±changed
+50−7
claude-codechanged
errors
| `the source file is not valid UTF-8 text` / `the source file is not valid UTF-16 text`
| `the source file has the replacement character U+FFFD`
| `Couldn't start a background session (working directory no longer exists or is not accessible: ...)`
```
+50 −7 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
cloud-environments
±changed
+4−4
claude-codechanged
cloud-environments
The agent proxy attaches a credential to a request when the request's host matches one you listed on that credential. Sessions can reach those hosts even when the environment's [ne
The agent proxy attaches a credential to a request when the request's host matches one you listed on that credential. Sessions can reach those hosts even when the environment's [ne
* **The Anthropic API and public package registries**: requests to `api.anthropic.com`, `registry.npmjs.org`, `jsr.io`, `npm.jsr.io`, `pypi.org`, `files.pythonhosted.org`, `index.c
* **The Anthropic API and public package registries**: `api.anthropic.com`, `registry.npmjs.org`, `jsr.io`, `npm.jsr.io`, `pypi.org`, `files.pythonhosted.org`, `index.crates.io`, a
* The hosts you listed on the environment's [API credentials](#add-api-credentials), except the [hosts the agent proxy skips](#requests-that-never-get-the-credential)
+4 −4 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
cli-reference
±changed
+1−1
claude-codechanged
cli-reference
| `--effort` | Set the [effort level](/docs/en/model-config#adjust-effort-level) for the current session. Options: `low`, `medium`, `high`, `xh
| `--effort` | Set the [effort level](/docs/en/model-config#adjust-effort-level) for the current session. Options: `low`, `medium`, `high`, `xh
+1 −1 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
artifacts
±changed
+1−1
claude-codechanged
artifacts
| Source file types | The published file must be `.html`, `.htm`, or `.md`. Markdown files render as styled HTML.
| Source file types | The published file must be `.html`, `.htm`, or `.md`, and must decode as UTF-8, or as little-endian UTF-16 by its byte-order mark. Markdown files render as st
+1 −1 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-code
agent-sdk/typescript
±changed
+1−1
claude-codechanged
agent-sdk/typescript
`effortLevel` accepts an [effort level](/docs/en/model-config#adjust-effort-level) name. It also accepts `"ultracode"`, which runs the session at `xhigh` effort and turns on [ultra
`effortLevel` accepts an [effort level](/docs/en/model-config#adjust-effort-level) name. It also accepts `"ultracode"`, which requests `xhigh` effort with [ultracode](/docs/en/work
+1 −1 lines
Recorded here 10 Sep 2026 18:07 UTC
claude-docs
claude-tag/admins/federated-access/troubleshooting
+new page
+345−0
claude-docsnew page
claude-tag/admins/federated-access/troubleshooting
No line of the body moved in this read.
+345 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/token-reference
+new page
+153−0
claude-docsnew page
claude-tag/admins/federated-access/token-reference
No line of the body moved in this read.
+153 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/overview
+new page
+77−0
claude-docsnew page
claude-tag/admins/federated-access/overview
No line of the body moved in this read.
+77 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/limits
+new page
+97−0
claude-docsnew page
claude-tag/admins/federated-access/limits
No line of the body moved in this read.
+97 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/gcp
+new page
+159−0
claude-docsnew page
claude-tag/admins/federated-access/gcp
No line of the body moved in this read.
+159 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/connect-a-gateway
+new page
+120−0
claude-docsnew page
claude-tag/admins/federated-access/connect-a-gateway
No line of the body moved in this read.
+120 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/aws
+new page
+150−0
claude-docsnew page
claude-tag/admins/federated-access/aws
No line of the body moved in this read.
+150 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/authorization-server
+new page
+137−0
claude-docsnew page
claude-tag/admins/federated-access/authorization-server
No line of the body moved in this read.
+137 −0 lines
Recorded here 10 Sep 2026 17:37 UTC
claude-docs
claude-tag/admins/federated-access/overview
±changed
+13−13
claude-docschanged
claude-tag/admins/federated-access/overview
| Connection type | Who accepts the token
| :------------------------ | :-----------------------------------------------------------------------------------------------------------------------------------------------------
| **Gateway** | A service you run. It verifies the token, maps the agent to permissions, and forwards the request to your internal systems with credentials you hold.
| **AWS role** | AWS, through an IAM OIDC identity provider. AWS issues temporary credentials for a role whose trust policy names Anthropic's issuer and your organizat
| **Google Cloud identity** | Google Cloud, through a workload identity pool. Google issues an access token for the federated identity, acting as a service account if you name one.
+13 −13 lines
Upstream edited 10 Sep 2026 17:36 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
third-party/claude-desktop/data-storage
±changed
+1−1
claude-docschanged
third-party/claude-desktop/data-storage
| `vm_bundles/` | Cached copy of the Cowork sandbox VM bundle.
| `vm_bundles/` | Cached copy of the VM workspace bundle that Cowork sessions run in, plus the data disk the app creates to hold thos
+1 −1 lines
Recorded here 10 Sep 2026 17:08 UTC
claude-docs
claude-tag/users/use-cases/triage-requests
±changed
+24−5
claude-docschanged
claude-tag/users/use-cases/triage-requests
> Claude Tag triages a Slack request channel with two setup messages. See in-thread answers, duplicate flags, owner routing, weekly theme rollups, and optional ticket filing.
> Claude Tag triages a Slack request channel with two setup messages. See in-thread answers, duplicate flags, owner routing, weekly theme rollups, optional ticket filing, and runbo
| Connection | Examples | Why it matters here |
| :------------- | :----------- | :-------------------------------------- |
| None | — | Works on Slack content alone |
+24 −5 lines
Recorded here 10 Sep 2026 17:08 UTC
claude-code
monitoring-usage
±changed
+1−1
claude-codechanged
monitoring-usage
* When authenticated via OAuth, `user.email` is included in telemetry attributes. If this is a concern for your organization, work with your telemetry backend to filter or redact t
* When authenticated via OAuth, `user.email` is included in telemetry attributes, sent only to the OTel endpoint you configure, never to Anthropic. If this is a concern for your or
+1 −1 lines
Recorded here 10 Sep 2026 17:07 UTC
claude-docs
claude-tag/users/use-cases/triage-requests
±changed
+10−10
claude-docschanged
claude-tag/users/use-cases/triage-requests
Check that the channel has the connections below. Ask `@Claude what can you access from this channel?` to check; an admin can [add a connection](/docs/claude-tag/admins/add-connect
Check that the channel has the connections below. Ask `@Claude what can you access from this channel?` to check; an admin can [add a connection](/docs/claude-tag/admins/add-connect
"Remember for this channel" saves the role to [channel memory](/docs/claude-tag/users/memory), so it applies to everyone's threads, not just yours. Tell requesters to include `@Cla
"Remember for this channel" saves the role to [channel memory](/docs/claude-tag/users/memory.md), so it applies to everyone's threads, not just yours. Tell requesters to include `@
"Including posts that didn't tag you" turns the recap into a sweep, so requests that arrived without a mention are still caught. To list or cancel scheduled work later, see [Manage
+10 −10 lines
Upstream edited 10 Sep 2026 16:53 UTC Recorded here 22 Oct 2026 18:07 UTC
claude-docs
claude-tag/users/use-cases/triage-requests
±changed
+21−25
claude-docschanged
claude-tag/users/use-cases/triage-requests
export const BetaNote = () => <Info>Claude Tag is in public beta. Features and behavior described here may change before general availability.</Info>;
<BetaNote />
> [!NOTE]
>
+21 −25 lines
Upstream edited 10 Sep 2026 16:53 UTC Recorded here 20 Oct 2026 18:07 UTC
claude-docs
claude-tag/users/use-cases/triage-requests
±changed
+4−4
claude-docschanged
claude-tag/users/use-cases/triage-requests
| Connection | Examples | Why it matters here |
| :----------------- | :------------------- | :---------------------------------------------------------------------------------------------------------------- |
| None | — | Works on Slack content alone |
| Connection | Examples | Why it matters here |
| :- | :- | :- |
+4 −4 lines
Upstream edited 10 Sep 2026 16:53 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
cowork/changelog
±changed
+1−1
claude-docschanged
cowork/changelog
A Windows update released September 8, 2026 (including KB5124008) stops Cowork from reaching your files when it runs on your Windows PC, so tasks fail or the workspace does not s
A Windows update released September 8, 2026 (including KB5124008) stops Cowork from reaching your files when it runs on your Windows PC, so tasks fail or the workspace does not s
+1 −1 lines
Recorded here 10 Sep 2026 16:07 UTC
claude-docs
claude-tag/admins/set-spend-limit
±changed
+3−3
claude-docschanged
claude-tag/admins/set-spend-limit
Channel work can't be attributed to individual users. It bills to your organization's usage balance, not to any user's seat, and often has no single requesting user (several people
The usage page at [`claude.ai/admin-settings/usage/claude-tag`](https://claude.ai/admin-settings/usage/claude-tag) shows spend broken down by channel.
In claude.ai you see spend per channel, not per user. The usage page at [`claude.ai/admin-settings/usage/claude-tag`](https://claude.ai/admin-settings/usage/claude-tag) shows each
+3 −3 lines
Recorded here 10 Sep 2026 16:07 UTC
claude-docs
claude-tag/admins/customize
±changed
+1−1
claude-docschanged
claude-tag/admins/customize
The Configure page also shows the channel's resolved access. Its **Tools and access** tab lists the channel's resolved connections and any allowed domains. Members can see those li
The Configure page also shows the channel's resolved access. Its **Tools and access** tab lists the channel's resolved connections and any allowed domains. Members can see those li
+1 −1 lines
Recorded here 10 Sep 2026 16:07 UTC
claude-docs
claude-tag/admins/attribute-costs
+new page
+48−0
claude-docsnew page
claude-tag/admins/attribute-costs
No line of the body moved in this read.
+48 −0 lines
Recorded here 10 Sep 2026 16:07 UTC
claude-docs
claude-tag/admins/add-connections
±changed
+11−5
claude-docschanged
claude-tag/admins/add-connections
A connection grants access; a plugin teaches Claude how to use it well. A plugin is a bundle of skills, reusable instructions for working with a specific tool or following a specif
A Datadog API key, for example, makes the API reachable, and a Datadog plugin tells Claude which endpoints answer which questions. Sessions in covered channels pick up attached plu
Anthropic provides plugins for common tools, and you can add your own from a [skills repository](/docs/claude-tag/admins/skills-repo). To give Claude organization-wide skills, bund
+11 −5 lines
Recorded here 10 Sep 2026 16:07 UTC
claude-docs
cowork/changelog
±changed
+4−0
claude-docschanged
cowork/changelog
<Update label="Known issue: Cowork on Windows" description="2026-09-10">
A Windows update released September 8, 2026 (including KB5124008) stops Cowork from reaching your files when it runs on your Windows PC, so tasks fail or the workspace does not s
</Update>
+4 −0 lines
Recorded here 10 Sep 2026 15:37 UTC
claude-code
workflows
±changed
+1−1
claude-codechanged
workflows
An `agent()` call resolves to `null` if you stop it mid-run or it hits an unrecoverable API error. `pipeline()` keeps that `null` in the results array, which is why the example end
An `agent()` call resolves to `null` if you stop it mid-run or it hits an unrecoverable API error. In [auto mode](/docs/en/permission-modes#eliminate-prompts-with-auto-mode), the c
+1 −1 lines
Recorded here 10 Sep 2026 14:37 UTC
claude-code
plugin-marketplaces
±changed
+1−1
claude-codechanged
plugin-marketplaces
Paths resolve relative to the marketplace root, which is the directory containing `.claude-plugin/`. In the example above, `./plugins/my-plugin` points to `<repo>/plugins/my-plugin
Paths resolve relative to the marketplace root, which is the directory containing `.claude-plugin/`. In the example above, `./plugins/my-plugin` points to `<repo>/plugins/my-plugin
+1 −1 lines
Recorded here 10 Sep 2026 14:37 UTC
claude-code
model-config
±changed
+2−0
claude-codechanged
model-config
On Fable 5, Opus 4.8, and Opus 4.7, frontmatter effort also applies while the [hold on the model's default effort](#adjust-effort-level) is in effect. Before v2.1.267, the hold too
+2 −0 lines
Recorded here 10 Sep 2026 14:37 UTC
claude-code
fullscreen
±changed
+2−0
claude-codechanged
fullscreen
While you're scrolled up, a dim header row at the top of the conversation shows the most recent prompt that has scrolled above the view. Click the row to jump to that prompt.
+2 −0 lines
Recorded here 10 Sep 2026 14:37 UTC
claude-docs
third-party/claude-desktop/extensions
±changed
+9−9
claude-docschanged
third-party/claude-desktop/extensions
| File | Purpose
| ---------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------
| `.claude-plugin/plugin.json` | **Required.** Plugin manifest (name, description, version). Directories without this file are ignored.
| `version.json` | `{"version": "1.2.3"}`. When this string changes, Claude Desktop re-syncs the plugin on next launch. Any string change triggers re-sync (there's no
| `.mcp.json` | MCP servers bundled with this plugin. A JSON object keyed by server name: `{"mcpServers": {"<name>": {"type": "http", "url": "...", "oauth": true}}
+9 −9 lines
Recorded here 10 Sep 2026 13:07 UTC
claude-code
settings-reference
±changed
+6−2
claude-codechanged
settings-reference
An entry written with the trailing dot that marks a fully qualified domain name, such as `example.com.`, blocks the same connections as `example.com`.
When server-managed settings deliver the policy at launch, they take precedence over the helper's source and the helper doesn't run. If a later settings fetch reports the server-ma
When server-managed settings deliver the policy at launch, they take precedence over the helper's source and the helper doesn't run.
+6 −2 lines
Recorded here 10 Sep 2026 04:37 UTC
claude-code
server-managed-settings
±changed
+6−3
claude-codechanged
server-managed-settings
Every other key in the cached `env` block applies at startup. Once the server confirms the payload, and you approve it if it needs [security approval](#security-approval-dialogs),
Every other key in the cached `env` block applies at startup. Once the server confirms the payload, and you approve it if it needs [security approval](#security-approval-dialogs),
Clients signed in through a [Claude apps gateway](#platform-availability) wait for the startup fetch whether or not you set this. If the gateway answers an attended interactive lau
Clients signed in through a [Claude apps gateway](#platform-availability) wait for the startup fetch whether or not you set this, and handle a failed fetch as follows:
+6 −3 lines
Recorded here 10 Sep 2026 04:37 UTC
claude-code
managed-settings
±changed
+2−2
claude-codechanged
managed-settings
* A new or changed [`policyHelper`](/docs/en/settings-reference#policyhelper) entry takes effect at the next launch, except that a helper shadowed at launch by server-managed set
* A new or changed [`policyHelper`](/docs/en/settings-reference#policyhelper) entry takes effect at the next launch. If server-managed settings shadow the helper at that launch,
This table shows how Claude Code combines each kind of key under `"merge"`. The [`managedSourcesBehavior` entry](/docs/en/settings-reference#managedsourcesbehavior) names every key
This table shows how Claude Code combines each kind of key under `"merge"`. The [`managedSourcesBehavior` entry](/docs/en/settings-reference#managedsourcesbehavior) names every key
+2 −2 lines
Recorded here 10 Sep 2026 04:37 UTC
claude-code
self-hosted-environments-reference
±changed
+1−1
claude-codechanged
self-hosted-environments-reference
| `--use-anthropic-git-proxy` | `CLAUDE_RUNNER_USE_GIT_PROXY=1` | off | Clone via Anthropic's git proxy instead of custome
| `--use-anthropic-git-proxy` | `CLAUDE_RUNNER_USE_GIT_PROXY=1` | off | Clone via the [Anthropic git proxy](/docs/en/self-
+1 −1 lines
Recorded here 10 Sep 2026 04:07 UTC
claude-code
self-hosted-environments-deploy
±changed
+2−0
claude-codechanged
self-hosted-environments-deploy
The runner also reports the opt-in to Anthropic when it registers, printing `Registering as opted in to Anthropic-managed git (--use-anthropic-git-proxy)` at startup. Each session
+2 −0 lines
Recorded here 10 Sep 2026 04:07 UTC
9 Sep 2026
129 pages
+2461 −1114 lines
api
claude_api_primer
±changed
+1−1
apichanged
claude_api_primer
| `name` | The name of the tool. Must match the regex `^[a-zA-Z0-9_-]{1,64}$`. |
| `name` | The name of the tool. Must match the regex `^[a-zA-Z0-9_-]{1,128}$`. |
+1 −1 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/completions/create
±changed
+0−2
apichanged
api/completions/create
- `Completion object`
+0 −2 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/completions
±changed
+0−2
apichanged
api/completions
- `Completion object`
+0 −2 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions/threads/events/stream
±changed
+0−4
apichanged
api/beta/sessions/threads/events/stream
- `BetaManagedAgentsStreamSessionThreadEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in a single thread's stream.
+0 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions/threads/events
±changed
+0−4
apichanged
api/beta/sessions/threads/events
- `BetaManagedAgentsStreamSessionThreadEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in a single thread's stream.
+0 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions/threads
±changed
+0−4
apichanged
api/beta/sessions/threads
- `BetaManagedAgentsStreamSessionThreadEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in a single thread's stream.
+0 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions/events/stream
±changed
+0−4
apichanged
api/beta/sessions/events/stream
- `BetaManagedAgentsStreamSessionEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in the session stream.
+0 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions/events
±changed
+0−4
apichanged
api/beta/sessions/events
- `BetaManagedAgentsStreamSessionEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in the session stream.
+0 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta/sessions
±changed
+0−8
apichanged
api/beta/sessions
- `BetaManagedAgentsStreamSessionEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in the session stream.
- `BetaManagedAgentsStreamSessionThreadEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
+0 −8 lines
Recorded here 9 Sep 2026 23:07 UTC
api
api/beta
±changed
+0−8
apichanged
api/beta
- `BetaManagedAgentsStreamSessionEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
Server-sent event in the session stream.
- `BetaManagedAgentsStreamSessionThreadEvents = BetaManagedAgentsUserMessageEvent or BetaManagedAgentsUserInterruptEvent or BetaManagedAgentsUserToolConfirmationEvent or 34 more`
+0 −8 lines
Recorded here 9 Sep 2026 23:07 UTC
api
agents-and-tools/tool-use/define-tools
±changed
+1−1
apichanged
agents-and-tools/tool-use/define-tools
| `name` | The name of the tool. Must match the regex `^[a-zA-Z0-9_-]{1,64}$`.
| `name` | The name of the tool. Must match the regex `^[a-zA-Z0-9_-]{1,128}$`.
+1 −1 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
settings-reference
±changed
+7−3
claude-codechanged
settings-reference
When a managed source on the machine sets `"gateway"`, Claude Code doesn't use a leftover login, API key, or `apiKeyHelper` credential. See [Administrator policy requires a Cloud g
Set the gateway URL the `/login` Cloud gateway screen connects to, so people reach your [cloud gateway](/docs/en/claude-apps-gateway) without typing its address. The screen has no
Set the gateway URL the `/login` Cloud gateway screen connects to, so people reach your [cloud gateway](/docs/en/claude-apps-gateway) without typing its address. The screen has no
+7 −3 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
llm-gateway-rollout
±changed
+1−1
claude-codechanged
llm-gateway-rollout
Do not include `forceLoginMethod` or `forceLoginOrgUUID` in managed settings alongside a gateway credential. Either key, with any value, blocks `ANTHROPIC_API_KEY`, `ANTHROPIC_AUTH
Don't include `forceLoginMethod` or `forceLoginOrgUUID` in managed settings alongside a gateway credential. Either key, with any value, blocks `ANTHROPIC_API_KEY`, `ANTHROPIC_AUTH_
+1 −1 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
llm-gateway-protocol
±changed
+12−3
claude-codechanged
llm-gateway-protocol
Token-counting endpoints are the only optional ones: when they're absent, Claude Code falls back to counting context usage through the inference endpoint instead. Inference request
Token-counting endpoints are the only optional ones: when they're absent, Claude Code falls back to a character-based estimate of context usage.
Match on the path, not the full URL:
+12 −3 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
errors
±changed
+34−1
claude-codechanged
errors
| `Not signed in to the Cloud gateway — run /login.`
| `Administrator policy requires a Cloud gateway sign-in on this machine`
| `Claude Code may not be enabled for your organization` after a Claude apps gateway sign-in
* If `/status` shows an `API key` row, an approved [`ANTHROPIC_API_KEY`](/docs/en/authentication#authentication-precedence) is the active credential and takes precedence over your
* If `/status` shows an `API key` row that isn't marked as not in use, an approved [`ANTHROPIC_API_KEY`](/docs/en/authentication#authentication-precedence) is the active credential
+34 −1 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
claude-apps-gateway-on-aws
±changed
+4−2
claude-codechanged
claude-apps-gateway-on-aws
* **IAM role** with `bedrock:InvokeModel` and `bedrock:InvokeModelWithResponseStream`, attached as the ECS task role or bound via IAM Roles for Service Accounts (IRSA) on EKS
* **IAM role** with `bedrock:InvokeModel`, `bedrock:InvokeModelWithResponseStream`, and `bedrock:CountTokens`, attached as the ECS task role or bound via IAM Roles for Service Acco
"Action": ["bedrock:InvokeModel", "bedrock:InvokeModelWithResponseStream"],
"Action": ["bedrock:InvokeModel", "bedrock:InvokeModelWithResponseStream", "bedrock:CountTokens"],
+4 −2 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
claude-apps-gateway-deploy
±changed
+40−31
claude-codechanged
claude-apps-gateway-deploy
Don't redirect requests to the gateway's device-authorization and token endpoints. Claude Code doesn't follow redirects on those requests, so an ingress rule that redirects them, s
Don't redirect requests to the gateway's device-authorization and token endpoints, for example with an HTTP-to-HTTPS or host-canonicalization rewrite at the ingress. Claude Code do
Once the gateway is serving, push `forceLoginMethod`, `forceLoginGatewayUrl`, and `parentSettingsBehavior: "merge"` to each developer's machine through managed settings, via MDM or
Once the gateway is serving, push `forceLoginMethod`, `forceLoginGatewayUrl`, and `parentSettingsBehavior: "merge"` to each developer's machine through managed settings, via MDM or
+40 −31 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
claude-apps-gateway
±changed
+6−4
claude-codechanged
claude-apps-gateway
The Amazon Bedrock upstream needs an AWS principal with `bedrock:InvokeModel` and `bedrock:InvokeModelWithResponseStream` on both the `inference-profile/us.anthropic.*` ARNs
The Amazon Bedrock upstream needs an AWS principal with `bedrock:InvokeModel` and `bedrock:InvokeModelWithResponseStream` on both the `inference-profile/us.anthropic.*` ARNs
Supply the credential with IRSA on EKS, an ECS task role, or an EC2 instance profile rather than static keys. The [`upstreams` reference](/docs/en/claude-apps-gateway-config#
The developer presses Enter to connect. The [first-connect TLS fingerprint prompt](#connect-developers) still appears.
+6 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
authentication
±changed
+8−4
claude-codechanged
authentication
If you have an active Claude subscription but also have `ANTHROPIC_API_KEY` set in your environment, the API key takes precedence once approved. This can cause authentication failu
If your machine's [managed settings](/docs/en/managed-settings) set [`forceLoginMethod`](/docs/en/settings-reference#forceloginmethod) to `"gateway"` or set [`forceLoginGatewayUrl`
If you have an active Claude subscription but also have `ANTHROPIC_API_KEY` set in your environment, Claude Code uses the API key once you approve it. This can cause authentication
+8 −4 lines
Recorded here 9 Sep 2026 23:07 UTC
claude-code
sub-agents
±changed
+1−1
claude-codechanged
sub-agents
This works with built-in and custom subagents, and the choice persists when you resume the session: Claude Code restores the agent's system prompt, tool restrictions, and model alo
This works with built-in and custom subagents, and the choice persists when you resume the session: Claude Code restores the agent's tool restrictions and model along with the conv
+1 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
settings-reference
±changed
+3−3
claude-codechanged
settings-reference
Select an [output style](/docs/en/output-styles) by name. An output style is a saved set of instructions that Claude Code adds to the system prompt to change Claude's role, tone, a
If you change this key during a session, Claude uses the new style starting with your next message. That message rebuilds the [prompt cache](/docs/en/prompt-caching#changing-output
Select an [output style](/docs/en/output-styles) by name. An output style is a saved set of instructions that changes Claude's role, tone, and output format, such as the built-in E
+3 −3 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
sessions
±changed
+2−2
claude-codechanged
sessions
* Agent: a session started with [`--agent`](/docs/en/sub-agents#invoke-subagents-explicitly) or the `agent` setting continues as that agent, keeping its system prompt, tool restric
* Agent: a session started with [`--agent`](/docs/en/sub-agents#invoke-subagents-explicitly) or the `agent` setting continues as that agent, keeping its tool restrictions and model
Not every configuration flag from the original launch is restored. If the session depended on `--mcp-config`, `--settings`, `--plugin-dir`, `--fallback-model`, or directories added
Not every configuration flag from the original launch is restored. If the session depended on `--mcp-config`, `--settings`, `--plugin-dir`, `--fallback-model`, or directories added
+2 −2 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
sandboxing
±changed
+3−0
claude-codechanged
sandboxing
* **0-byte read-only files appear at `.claude` settings paths, and "Yes, and don't ask again" doesn't save**: on Linux and WSL2, the sandbox holds a write denial on a file that doe
Run `claude doctor` to list the leftover placeholder files. The [`Stale sandbox mask files left by a killed session`](/docs/en/errors#stale-sandbox-mask-files-left-by-a-killed-se
+3 −0 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
prompt-caching
±changed
+8−6
claude-codechanged
prompt-caching
| Layer | Content | Changes when |
| --------------- | ------------------------------------------------- | ----------------------------------------------------------------------------------------------- |
| System prompt | Core instructions, tool definitions, output style | The set of loaded tool definitions changes, you switch output style, or Claude Code is upgraded |
| Project context | CLAUDE.md, auto memory, unscoped rules | Session starts, or after `/clear` or `/compact` |
| Conversation | Your messages, Claude's responses, tool results | Every turn |
+8 −6 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
plugins-reference
±changed
+3−1
claude-codechanged
plugins-reference
| `displayName` | string | Human-readable name shown in the `/plugin` picker and other UI surfaces. Falls back to `name` when omitted. Unlike `name`, may contain spaces and any
| `displayName` | string | Human-readable name shown in the `/plugin` picker and other UI surfaces. For a marketplace-installed plugin, a `displayName` on the [marketplace entr
On macOS and Linux, Claude Code also rejects a component path that contains a backslash anywhere in it, even when the path stays inside the plugin. Components declared with backsla
+3 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
plugin-marketplaces
±changed
+12−5
claude-codechanged
plugin-marketplaces
**Reserved names**: the following marketplace names are reserved for official Anthropic use and can't be used by third-party marketplaces: `claude-code-marketplace`, `claude-code
Claude Code re-checks reserved names every time it loads a marketplace, not only when you add one. A marketplace that was registered under one of these names before the name beca
**Reserved names**: the following marketplace names are reserved for official Anthropic use and can't be used by third-party marketplaces: `claude-code-marketplace`, `claude-code
+12 −5 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
output-styles
±changed
+21−17
claude-codechanged
output-styles
Output styles change how Claude responds, not what Claude knows. They modify the system prompt to set role, tone, and output format. Use one when you keep re-prompting for the same
Output styles change how Claude responds, not what Claude knows. They set Claude's role, tone, and output format for every response. Use one when you keep re-prompting for the same
A custom output style adds your instructions to the system prompt and lets you choose whether to keep Claude Code's built-in software engineering instructions. Keep them when you'r
A custom output style gives Claude your own instructions and lets you choose whether to keep Claude Code's built-in software engineering instructions. Keep them when you're changin
Claude Code's **Default** output style is the existing system prompt, designed to help you complete software engineering tasks efficiently.
+21 −17 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
network-config
±changed
+1−0
claude-codechanged
network-config
| `*-review.googlesource.com` | Gerrit change lookup on `googlesource.com` checkouts. When a Claude Desktop Code tab session starts or resumes on a [trusted](/docs/en/perm
+1 −0 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
memory
±changed
+1−1
claude-codechanged
memory
For instructions you want at the system prompt level, use [`--append-system-prompt`](/docs/en/cli-reference#system-prompt-flags). This must be passed every invocation, so it's bett
For instructions you want at the system prompt level, use [`--append-system-prompt`](/docs/en/cli-reference#system-prompt-flags). You pass it at launch, so it's better suited to sc
+1 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
glossary
±changed
+1−1
claude-codechanged
glossary
A configuration that modifies Claude's system prompt to change response behavior, tone, or format. Unlike [CLAUDE.md](#claude-md), which Claude Code delivers as a user message afte
A configuration that changes the instructions Claude Code gives Claude, to set response behavior, tone, or format. Unlike [CLAUDE.md](#claude-md), which adds project context alongs
+1 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
errors
±changed
+118−2
claude-codechanged
errors
| `couldn't be resolved to a real location, so its skills, commands, and agents weren't loaded`
| `path could not be checked`
| `its marketplace entry path does not stay inside the marketplace directory`
| `Plugin source path refused`
| `Command killed: its output file was replaced or could no longer be verified`
+118 −2 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
env-vars
±changed
+1−0
claude-codechanged
env-vars
* Have a resumed conversation [keep its recorded system prompt](/docs/en/cli-reference#system-prompt-flags-in-resumed-conversations); Claude Code rebuilds the prompt on every reque
+1 −0 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
discover-plugins
±changed
+1−1
claude-codechanged
discover-plugins
URL-based marketplaces have some limitations compared to Git-based marketplaces. If you encounter "path not found" errors when installing plugins, see [Troubleshooting](/docs/en/
URL-based marketplaces have some limitations compared to Git-based marketplaces. If plugin installs from a URL-based marketplace fail, see [Troubleshooting](/docs/en/plugin-marke
+1 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
context-window
±changed
+2−2
claude-codechanged
context-window
* **Before you type anything**: CLAUDE.md, auto memory, MCP tool names, and skill descriptions all load into context. Your own setup may add more here, like an [output style](/docs
* **Before you type anything**: CLAUDE.md, auto memory, MCP tool names, and skill descriptions all load into context. Your own setup may add more here, like an [output style](/docs
| System prompt and output style | Unchanged; not part of message history
| System prompt and output style | Both still apply
+2 −2 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
cli-reference
±changed
+18−8
claude-codechanged
cli-reference
| `--system-prompt-snapshot` | Pass `off` to rebuild the system prompt on every request instead of reusing the prompt [recorded on the conversation's first req
Claude Code provides four flags for customizing the system prompt. All four work in both interactive and non-interactive modes.
Claude Code provides five flags for customizing the system prompt. Four set its text, and with `--system-prompt-snapshot` you control whether a conversation keeps the text it start
| Flag | Behavior | Example |
| :---------------------------- | :------------------------------------------ | :------------------------------------------------------ |
+18 −8 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
claude-directory
±changed
+8−8
claude-codechanged
claude-directory
contains: [<><A href="/docs/en/permissions">permissions</A>: allow, deny, or prompt before Claude uses specific tools or commands</>, <><A href="/docs/en/hooks">hooks</A>
contains: [<><A href="/docs/en/permissions">permissions</A>: allow, deny, or prompt before Claude uses specific tools or commands</>, <><A href="/docs/en/hooks">hooks</A>
when: 'Files read at startup; the style you select with outputStyle is added to the system prompt every turn',
when: 'Files read at startup; the style you select with outputStyle applies to every response',
oneLiner: 'Custom system-prompt sections that adjust how Claude works',
+8 −8 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
claude-code-on-the-web
±changed
+2−2
claude-codechanged
claude-code-on-the-web
Claude Code computes these diffs, including the per-file diffs shown as Claude edits, from raw git blob content, so diff drivers and `textconv` filters configured in the repository
Claude Code computes these diffs, including the per-file diffs shown as Claude edits, from raw git blob content, so diff drivers and `textconv` filters configured in the repository
Cloud sessions stop after a period of inactivity and the session's VM is reclaimed. On the web, the session is marked expired in the session list.
Cloud sessions stop after a period of inactivity and the session's VM is reclaimed. A session counts as inactive while it waits for you to approve an [MCP connector](/docs/en/cloud
+2 −2 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
agent-view
±changed
+1−1
claude-codechanged
agent-view
When the backgrounded session later resumes or restarts, Claude Code restores the agent's system prompt and tool restrictions. It searches the session's own directory for the agent
When the backgrounded session later resumes or restarts, Claude Code restores the agent and its tool restrictions; for its system prompt, see [System prompt flags in resumed conver
+1 −1 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
agent-sdk/typescript
±changed
+67−67
claude-codechanged
agent-sdk/typescript
| Property | Type | Default
| :-------------------------------- | :------------------------------------------------------------------------------------------------------------------- | :----------------------
| `abortController` | `AbortController` | `new AbortController()`
| `additionalDirectories` | `string[]` | `[]`
| `agent` | `string` | `undefined`
+67 −67 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-code
agent-sdk/modifying-system-prompts
±changed
+10−2
claude-codechanged
agent-sdk/modifying-system-prompts
Output styles, `append`, and a custom prompt string each change the system prompt directly. CLAUDE.md takes a different path: the SDK reads it and injects its content into the conv
`append` and a custom prompt string each change the system prompt directly, and an output style changes the instructions Claude Code gives Claude for every response. CLAUDE.md take
Output styles are saved configurations that modify Claude's system prompt. They're stored as markdown files and can be reused across sessions and projects.
Output styles are saved sets of instructions that change Claude's role, tone, and output format. They're stored as markdown files and can be reused across sessions and projects.
+10 −2 lines
Recorded here 9 Sep 2026 22:37 UTC
claude-docs
government/security/security-and-data-handling
±changed
+17−9
claude-docschanged
government/security/security-and-data-handling
On macOS, those shell commands run inside an operating-system-level sandbox that the application builds from your organization's **Allowed network hosts** and **Allowed workspace f
The macOS sandbox also blocks connections to local Unix sockets. A command that talks to a local agent through a socket, such as git signing a commit with a key held in an SSH agen
Exempting git commands such as `git commit` with `sandbox.excludedCommands` is not a safe way around the socket restriction. A sandboxed command can still change files that git run
+17 −9 lines
Recorded here 9 Sep 2026 21:07 UTC
Nothing on this page matches that. Press Enter to search every page of the ledger.