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
808pages tracked
757 changed this week · last read 22 Oct 2026
Claude Code CLI
228pages tracked
152 changed this week · last read 22 Oct 2026
Claude Documentation
279pages tracked
76 changed this week · last read 22 Oct 2026
Model Context Protocol
357pages tracked
163 changed this week · last read 21 Oct 2026
100 on this page
5,531 of the ledger's changes in claude-code. Show every change
ReadSourcePageWhatLines
8 Sep 2026
34 pages
+247 −193 lines
claude-code
sessions
±changed
+4−0
claude-codechanged
sessions
### Delete session data
Transcripts age out under the [retention sweep rules](/docs/en/claude-directory#cleaned-up-automatically). To delete a project's transcripts and related state sooner, run [`claude
+4 −0 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
server-managed-settings
±changed
+1−1
claude-codechanged
server-managed-settings
* [Settings reference](/docs/en/settings-reference): every settings key
* [All settings](/docs/en/settings-reference): every settings key
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
sandboxing
±changed
+1−1
claude-codechanged
sandboxing
* [Settings reference](/docs/en/settings-reference): every settings key
* [All settings](/docs/en/settings-reference): every settings key
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
plugin-relevance
±changed
+1−1
claude-codechanged
plugin-relevance
* [Settings reference](/docs/en/settings-reference#pluginsuggestionmarketplaces): `pluginSuggestionMarketplaces` and `extraKnownMarketplaces`
* [All settings](/docs/en/settings-reference#pluginsuggestionmarketplaces): `pluginSuggestionMarketplaces` and `extraKnownMarketplaces`
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
permissions
±changed
+1−1
claude-codechanged
permissions
* [Settings reference](/docs/en/settings-reference#permission-settings): every settings key, including the permission keys
* [All settings](/docs/en/settings-reference#permission-settings): every settings key, including the permission keys
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
permission-modes
±changed
+5−1
claude-codechanged
permission-modes
Claude Code v2.1.261 and later also block these by default:
* Posting or writing a link to a public paste, diagram, or data-sharing service in a message, PR or issue text, a document, or anywhere else the link will be opened or fetched, whe
Pushing to any branch of the repository you're working in and creating a pull request that matches your request run without a prompt, unless the change would send secrets or sensit
+5 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
network-config
±changed
+1−1
claude-codechanged
network-config
* [Claude Code settings](/docs/en/settings)
* [Settings files and precedence](/docs/en/settings)
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
model-config
±changed
+22−8
claude-codechanged
model-config
`ANTHROPIC_CUSTOM_MODEL_OPTION_NAME` and `ANTHROPIC_CUSTOM_MODEL_OPTION_DESCRIPTION` are optional. If you omit the name, Claude Code uses the model ID; if you omit the description,
`ANTHROPIC_CUSTOM_MODEL_OPTION_NAME` and `ANTHROPIC_CUSTOM_MODEL_OPTION_DESCRIPTION` are optional:
* If you omit the name, the entry shows the model's name when Claude Code [recognizes the ID](#customize-pinned-model-display-and-capabilities), and the model ID otherwise.
* If you omit the description, Claude Code uses `Custom model (<model-id>)`.
+22 −8 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
managed-settings
±changed
+1−1
claude-codechanged
managed-settings
* [Settings reference](/docs/en/settings-reference): every key, with whether a managed source can set it
* [All settings](/docs/en/settings-reference): every key, with whether a managed source can set it
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
llm-gateway-protocol
±changed
+3−1
claude-codechanged
llm-gateway-protocol
The picker is the interactive model list that opens when a developer runs `/model` in Claude Code. Each discovered entry uses `display_name` as its name when the gateway sends one,
The picker is the interactive model list that opens when a developer runs `/model` in Claude Code. Each discovered entry uses `display_name` as its name when the gateway sends one
Discovery adds only models that the [`availableModels` managed setting](/docs/en/settings-reference#availablemodels) allows.
+3 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
hooks
±changed
+1−1
claude-codechanged
hooks
Claude Code runs ConfigChange hooks when a settings file, a managed policy file, or a skill file changes. For managed policy, it runs them only when `managed-settings.json` or a fi
Claude Code runs ConfigChange hooks when a settings file, a managed policy file, or a skill file changes. For managed policy, it runs them only when `managed-settings.json` or a fi
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
env-vars
±changed
+9−9
claude-codechanged
env-vars
| `ANTHROPIC_CUSTOM_MODEL_OPTION_NAME` | Display name for the custom model entry in the `/model` picker. Defaults to the model ID when not set
| `ANTHROPIC_CUSTOM_MODEL_OPTION_NAME` | Display name for the custom model entry in the `/model` picker. When not set, the entry shows the model's name if Claude
| `ANTHROPIC_DEFAULT_FABLE_MODEL_DESCRIPTION` | Display description for the pinned Fable model in the `/model` picker. Defaults to `Custom Fable model` when not set. Se
| `ANTHROPIC_DEFAULT_FABLE_MODEL_NAME` | Display name for the pinned Fable model in the `/model` picker. Defaults to the model ID when not set. See [Model config
| `ANTHROPIC_DEFAULT_FABLE_MODEL_DESCRIPTION` | Display description for the pinned Fable model in the `/model` picker. When not set, the row shows a default description
+9 −9 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
claude-code-on-the-web
±changed
+1−1
claude-codechanged
claude-code-on-the-web
* [Settings reference](/docs/en/settings-reference): all configuration options
* [All settings](/docs/en/settings-reference): all configuration options
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
auto-mode-config
±changed
+1−1
claude-codechanged
auto-mode-config
* [Settings reference](/docs/en/settings-reference#automode): every settings key, including `autoMode`
* [All settings](/docs/en/settings-reference#automode): every settings key, including `autoMode`
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
claude-code
admin-setup
±changed
+1−1
claude-codechanged
admin-setup
* [Settings reference](/docs/en/settings-reference): every setting key, with where it goes and an example
* [All settings](/docs/en/settings-reference): every setting key, with where it goes and an example
+1 −1 lines
Recorded here 8 Sep 2026 16:07 UTC
6 Sep 2026
1 page
+4 −0 lines
claude-code
changelog
±changed
+4−0
claude-codechanged
changelog
<Update label="2.1.263" description="September 6, 2026">
* Bug fixes and reliability improvements
</Update>
+4 −0 lines
Recorded here 6 Sep 2026 03:07 UTC
5 Sep 2026
37 pages
+824 −554 lines
claude-code
sub-agents
±changed
+1−1
claude-codechanged
sub-agents
In [non-interactive mode](/docs/en/headless), pass [`--append-subagent-system-prompt`](/docs/en/cli-reference#cli-flags) to append your text to the end of every subagent's system p
In [non-interactive mode](/docs/en/headless), pass [`--append-subagent-system-prompt`](/docs/en/cli-reference#cli-flags) to append your text to the end of every subagent's system p
+1 −1 lines
Recorded here 5 Sep 2026 14:37 UTC
claude-code
env-vars
±changed
+1−1
claude-codechanged
env-vars
| `CLAUDE_CODE_ENABLE_APPEND_SUBAGENT_PROMPT` | Set to `1` to enable appending extra text to the end of the system prompt of every [subagent](/docs/en/sub-agents) other
| `CLAUDE_CODE_ENABLE_APPEND_SUBAGENT_PROMPT` | Set to `1` to enable appending extra text to the end of the system prompt of every [subagent](/docs/en/sub-agents) other
+1 −1 lines
Recorded here 5 Sep 2026 14:37 UTC
claude-code
cli-reference
±changed
+1−0
claude-codechanged
cli-reference
| `--append-subagent-system-prompt-file` | Load text from a file and append it to [subagent](/docs/en/sub-agents) system prompts. An alternative to `--append-subagent-syst
+1 −0 lines
Recorded here 5 Sep 2026 14:37 UTC
claude-code
settings-reference
±changed
+5−3
claude-codechanged
settings-reference
* **Default**: unset, so every command Claude Code can sandbox runs sandboxed
* **Default**: unset, so no command is excluded
Let Claude retry a command outside the sandbox with the `dangerouslyDisableSandbox` parameter after the sandbox blocks it. Set it to `false` so Claude Code ignores that parameter c
Let Claude retry a command outside the sandbox with the `dangerouslyDisableSandbox` parameter after the sandbox blocks it. Set it to `false` so Claude Code ignores that parameter c
* `false`: Claude Code ignores that parameter, so every command runs sandboxed or appears in `excludedCommands`
+5 −3 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
settings-example
±changed
+3−3
claude-codechanged
settings-example
// Sandbox every command, refuse to start if the sandbox can't be set up, and
// never let a blocked command retry outside the sandbox; network limited to
// npm and GitHub, and users can't add domains
// Sandbox every command Claude runs, refuse to start if the sandbox can't be
// set up, and never let a blocked command retry outside the sandbox; network
+3 −3 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
sandboxing
±changed
+14−3
claude-codechanged
sandboxing
You can disable this escape hatch by setting `"allowUnsandboxedCommands": false` in your [sandbox settings](/docs/en/settings-reference#sandbox-settings). When disabled, which the
You can disable this escape hatch by setting `"allowUnsandboxedCommands": false` in your [sandbox settings](/docs/en/settings-reference#sandbox-settings). With the escape hatch dis
Strict sandbox mode applies to the commands Claude runs. Commands you type yourself at the [`!` shell-mode prompt](/docs/en/interactive-mode#shell-mode-with-prefix) run outside the
+14 −3 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
network-config
±changed
+1−1
claude-codechanged
network-config
* `CLAUDE_STREAM_FIRST_BYTE_TIMEOUT_MS` sets the first-byte deadline directly. Leave it unset and Claude Code derives the deadline from the byte-level watchdog's timeout or from an
* `CLAUDE_STREAM_FIRST_BYTE_TIMEOUT_MS` sets the first-byte deadline directly. Leave it unset and Claude Code uses the byte-level watchdog's timeout, so `CLAUDE_STREAM_IDLE_TIMEOUT
+1 −1 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
interactive-mode
±changed
+2−0
claude-codechanged
interactive-mode
In a regular interactive session, commands you type in shell mode run outside the [sandbox](/docs/en/sandboxing) even when you've enabled sandboxing, because the sandbox applies to
+2 −0 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
headless
±changed
+12−11
claude-codechanged
headless
| Field | Type | Description
| ---------------- | --------------- | ---------------------------------------------------------------------------------------------------------------------------------------------
| `type` | `"system"` | message type
| `subtype` | `"api_retry"` | identifies this as a retry event
| `attempt` | integer | current attempt number, starting at 1
+12 −11 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
errors
±changed
+17−17
claude-codechanged
errors
| [`API_TIMEOUT_MS`](/docs/en/env-vars) | 600000 | Per-request timeout in milliseconds. Raise it for slow networks or proxies. It also bounds the [first-byte
| [`API_TIMEOUT_MS`](/docs/en/env-vars) | 600000 | Per-request timeout in milliseconds. Raise it for slow networks or proxies. It also caps how long Claude Co
Claude Code sent a streaming request and the API returned no response headers within the deadline for the first byte, so Claude Code aborted the request instead of waiting for the
```text theme={null}
+17 −17 lines
Recorded here 5 Sep 2026 05:07 UTC
claude-code
worktrees
±changed
+1−1
claude-codechanged
worktrees
* **Command shape**: Claude Code blocks a Bash or Monitor command it can't verify stays inside the worktree, even when the command runs no git at all. Claude Code refuses shell con
* **Command shape**: Claude Code blocks a Bash or Monitor command when it can't verify from the command text that any git the command runs stays inside the worktree, for example wh
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
tools-reference
±changed
+1−1
claude-codechanged
tools-reference
This affects edit eligibility only, not permissions. [Read and Edit deny rules](/docs/en/permissions#tool-specific-permission-rules) also apply to file commands Claude Code recogni
Viewing a file with Bash affects edit eligibility only, not permissions. See [Read and Edit permission rules](/docs/en/permissions#read-and-edit) for which Bash commands your `Read
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
terminal-config
±changed
+1−1
claude-codechanged
terminal-config
When you delete with a word or line shortcut such as `Ctrl+W` or `Ctrl+K`, or with a vim delete through an `f`/`t` motion such as `df]`, and the deleted range reaches inside a plac
When you delete with a word or line shortcut such as `Ctrl+W` or `Ctrl+K`, or with a vim delete through an `f`/`t` motion such as `df]`, and the deleted range reaches inside a plac
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
sub-agents
±changed
+1−1
claude-codechanged
sub-agents
* It refuses a command whose shape it can't verify stays inside the worktree. This refusal applies even to a command that runs no git.
* It refuses a command when it can't verify from the command text that any git the command runs stays inside the worktree, for example when the command name is computed at runtime.
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
skills
±changed
+3−1
claude-codechanged
skills
Skills in nested `.claude/skills/` directories below your starting directory aren't loaded at startup. They load the first time Claude reads or edits a file inside that subdirector
Skills in nested `.claude/skills/` directories below your starting directory don't load at startup. They load the first time Claude reads or edits a file in the subdirectory that c
To load a subdirectory's skills before Claude reads or edits a file there, run `/add-dir` with that subdirectory's path. This requires Claude Code v2.1.257 or later.
+3 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
settings-reference
±changed
+11−17
claude-codechanged
settings-reference
| [`keybindingFlavor`](#keybindingflavor) | Make `Ctrl+W` [delete back to the previous whitespace](/docs/en/interactiv
| [`keybindingFlavor`](#keybindingflavor) | Deprecated and has no effect; the word-editing shortcuts always [follow re
List the tool uses Claude Code blocks. Use it for files that hold API keys, secrets, or environment values: Claude Code excludes matching files from file discovery and search resul
List the tool uses Claude Code blocks. Use it for files that hold API keys, secrets, or environment values: Claude Code excludes matching files from file discovery and search resul
Choose which convention `Ctrl+W` follows in the prompt input. Set it to `"readline"` to make `Ctrl+W` delete back to the previous whitespace, as Bash does, so a path or a `--flag=v
+11 −17 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
remote-control
±changed
+3−2
claude-codechanged
remote-control
A policy blocks Remote Control. Check these causes in order:
A policy blocks Remote Control, or Claude Code couldn't load your organization's policy on this machine and keeps Remote Control off in the meantime. Check these causes in order:
* **Otherwise, an Owner hasn't enabled it for your organization**: this form appears when you're signed in with an eligible claude.ai account but Remote Control is off, the default
* **The organization policy didn't load on this machine**: run `claude doctor` and read the `Organization policy` line. If the line shows the policy isn't loaded, that is what's ke
* **Otherwise, an Owner hasn't enabled it for your organization**: Remote Control is off by default on Team and Enterprise plans. An Owner can enable it at [claude.ai/admin-setting
+3 −2 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
plugins-reference
±changed
+3−1
claude-codechanged
plugins-reference
Claude Code doesn't let a plugin reference files outside its own directory. It rejects a component path that resolves outside the plugin root, such as `../shared-utils`, whether th
Claude Code doesn't let a plugin reference files outside its own directory. It rejects a component path that resolves outside the plugin root, whether the path is declared in `plug
When Claude Code rejects a path, it reports a [`path escapes plugin directory`](/docs/en/errors#path-escapes-plugin-directory) error and loads the plugin without that component.
+3 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
permissions
±changed
+20−5
claude-codechanged
permissions
Permission rules follow the format `Tool` or `Tool(specifier)`.
Permission rules follow the format `Tool` or `Tool(specifier)`. Parentheses inside the specifier are literal, so a command or path that contains them needs no escaping.
Deny and ask rules apply when any subcommand matches them, including a command nested inside a subshell, a command substitution, or a control-flow body such as a `for` loop. An ask
A `cd` into a path inside your working directory or an [additional directory](#working-directories) is also read-only, and a compound command like `cd packages/api && ls` runs with
+20 −5 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
mcp
±changed
+2−0
claude-codechanged
mcp
When you remove a remote server, Claude Code also deletes the OAuth tokens and client registration it stored for that server.
+2 −0 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
managed-settings
±changed
+2−0
claude-codechanged
managed-settings
To check whether these organization settings reached a given machine, run `claude doctor` there and read the `Organization policy` line, which says where Claude Code loaded the p
+2 −0 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
managed-mcp
±changed
+10−9
claude-codechanged
managed-mcp
Before loading a server, including one from `managed-mcp.json`, Claude Code runs the three checks below in order. In-process `type: "sdk"` servers, which the [app that started the
Before loading a server, including one from `managed-mcp.json`, Claude Code runs the three checks below in order. It runs them again when a user reconnects a server or turns a disa
| Restriction | What the user sees
| :------------------------------------------------------------------- | :---------------------------------------------------------------------------------------------------------
| `managed-mcp.json` is present and the user runs `claude mcp add` | `Cannot add MCP server: enterprise MCP configuration is active and has exclusive control over MCP servers`
+10 −9 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
llm-gateway-protocol
±changed
+20−10
claude-codechanged
llm-gateway-protocol
* Nonessential traffic is disabled, through [`CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC`](/docs/en/env-vars) or organization policy
Discovery still runs when [nonessential traffic is turned off](/docs/en/llm-gateway-connect#turn-off-traffic-outside-the-gateway-path), because the request goes only to your gatewa
Claude Code also sends any headers from `ANTHROPIC_CUSTOM_HEADERS`.
+20 −10 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
llm-gateway-connect
±changed
+4−2
claude-codechanged
llm-gateway-connect
Discovered models appear as additional `/model` entries labeled `From gateway`. To confirm discovery ran, start `claude --debug` and look for the `[gatewayDiscovery]` lines in the
Discovered models appear as additional `/model` entries. Each entry shows the description your gateway supplies for the model, or `From gateway` when it doesn't supply one.
To confirm discovery ran, start `claude --debug` and look for the `[gatewayDiscovery]` lines in the debug log at `~/.claude/debug/<session-id>.txt`. The first time discovery succee
* It turns off [gateway model discovery](#add-gateway-models-to-the-model-picker), even though discovery queries the gateway itself. Previously discovered models stay available fro
+4 −2 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
large-codebases
±changed
+1−1
claude-codechanged
large-codebases
Deny rules cover Claude's built-in file tools and recognized Bash file commands, including `cat`, `head`, `grep`, and `find`, when a denied path is passed as an argument. Claude Co
Deny rules cover Claude's built-in file tools. In Bash, they cover the file commands Claude Code recognizes, such as `cat`, `head`, `grep`, and `find`, when a denied path appears a
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
interactive-mode
±changed
+23−30
claude-codechanged
interactive-mode
| Shortcut | Description | Context
| :------------------------- | :----------------------------------- | :-------------------------------------------------------------------------------------------------------------
| `Ctrl+A` | Move cursor to start of current line | In multiline input, moves to the start of the current logical line
| `Ctrl+E` | Move cursor to end of current line | In multiline input, moves to the end of the current logical line
| `Ctrl+K` | Delete to end of line | Stores deleted text for pasting
+23 −30 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
hooks
±changed
+1−1
claude-codechanged
hooks
* You add a directory that is already a working directory; the add fails with an error
* You add a directory that is already a working directory or inside one
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
errors
±changed
+77−0
claude-codechanged
errors
| `is a network path, which cannot be added as a working directory`
| `MCP server <name> is blocked by enterprise managed policy`
| `Invalid permission rule "..." was skipped: Malformed Tool(content) rule`
Claude Code rejects both a path that points outside the plugin as written, such as `../shared-utils`, and a symlink that leads outside the plugin and isn't one the [marketplace sym
+77 −0 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
env-vars
±changed
+354−354
claude-codechanged
env-vars
No line of the body moved in this read.
+354 −354 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
deep-links
±changed
+1−1
claude-codechanged
deep-links
| `cwd` | Absolute path to use as the working directory. Network and UNC paths are rejected, and so are paths that contain invisible or bidirectional control characters.
| `cwd` | Absolute path to use as the working directory. Network and UNC paths are rejected, and so are paths that contain `..` segments or invisible or bidirectional control c
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
commands
±changed
+1−1
claude-codechanged
commands
| `/add-dir <path>` | Add a working directory for file access during the current session. Type a partial
| `/add-dir <path>` | Add a working directory for file access during the current session. Type a partial
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
cloud-environments
±changed
+4−1
claude-codechanged
cloud-environments
If sessions in the environment work with [artifacts](/docs/en/artifacts), include `*.frame.claudeusercontent.com` in your list. Claude Code fetches artifact content from that host.
If your organization uses [artifacts](/docs/en/artifacts#availability), you don't need `*.frame.claudeusercontent.com` in the list for sessions to read them. When the list leaves t
* **Sessions in this environment open another organization's public artifacts**: Claude Code fetches those from the host directly, so add it to this list.
* **You're configuring the local CLI or a self-hosted runner**: keep the host in that allowlist. See [network access requirements](/docs/en/network-config#network-access-requiremen
+4 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
cli-reference
±changed
+1−1
claude-codechanged
cli-reference
| `--add-dir` | Add additional working directories for Claude to read and edit files. Grants file access; most `.claude/` configuration is [not
| `--add-dir` | Add additional working directories for Claude to read and edit files. Grants file access; Claude Code [doesn't discover](/docs/e
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
agent-sdk/typescript
±changed
+13−9
claude-codechanged
agent-sdk/typescript
| `interrupt()` | Interrupts the query. Only available in streaming input mode. When the CLI advertises the `interrupt_receipt_v1` capability in [`SDKSyste
| `interrupt()` | Interrupts the query. Only available in streaming input mode. When the CLI advertises the `interrupt_receipt_v1` capability in [`SDKSyste
`still_queued` lists the UUIDs of user messages that survive the interrupt: messages still in the queue, plus any messages Claude Code had already taken off the queue for the next
`still_queued` lists the UUIDs of the user messages that were pending when the interrupt arrived: messages still in the queue, plus any messages Claude Code had already taken off t
+13 −9 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
accessibility
±changed
+1−1
claude-codechanged
accessibility
* Deleting a word with `Ctrl+W`, `Option+Delete` on macOS, or `Ctrl+Backspace` on Windows
* Deleting words with `Ctrl+W` or `Alt+D`, or with `Option+Delete` on macOS or `Ctrl+Backspace` on Windows
+1 −1 lines
Recorded here 5 Sep 2026 04:37 UTC
claude-code
tools-reference
±changed
+13−7
claude-codechanged
tools-reference
| Result | What Claude gets
| :------ | :-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
| Valid | Inline up to roughly 30,000 characters; past that, the path of a file saved to the session directory, truncated past 64 MiB, plus a short preview from the start, and Cl
| Failure | Inline up to roughly 10,000 characters; past that, a head-and-tail excerpt of that size cut from the read-back window, with no file path
| Result | What Claude gets
+13 −7 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
skills
±changed
+10−2
claude-codechanged
skills
* **Output size**: output past the Bash tool's inline ceiling arrives as a file path plus a short preview, not truncated text. [Output limits](/docs/en/tools-reference#output-limit
* **Output size**: output past the Bash tool's inline ceiling arrives as a file path plus a short preview, not truncated text. [Output limits](/docs/en/tools-reference#output-limit
### Find unused skills
+10 −2 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
settings-reference
±changed
+34−0
claude-codechanged
settings-reference
| [`bashOutputMaxChars`](#bashoutputmaxchars) | Set how much of a successful command's [output](/docs/en/tools-reference#o
| [`taskOutputMaxChars`](#taskoutputmaxchars) | Set how much of a [background task's](/docs/en/tools-reference#background-
### `bashOutputMaxChars`
Set how many characters of a successful Bash or PowerShell command's [output Claude receives inline](/docs/en/tools-reference#output-limits). When output passes the limit, Claude C
+34 −0 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
monitoring-usage
±changed
+1−1
claude-codechanged
monitoring-usage
* `"settings_unknowable"`: A settings file couldn't be read or parsed, so `cleanupPeriodDays` may be set to a value Claude Code can't see
* `"settings_unknowable"`: A settings file couldn't be read or parsed, so `cleanupPeriodDays` or `desktopSessionCleanupPeriodDays` may be set to a value Claude Code can't see
+1 −1 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
memory
±changed
+1−1
claude-codechanged
memory
Claude Code deletes old session transcripts after the [`cleanupPeriodDays`](/docs/en/settings-reference#cleanupperioddays) retention period, but excludes the files in the memory di
Claude Code deletes old session transcripts after the [`cleanupPeriodDays`](/docs/en/settings-reference#cleanupperioddays) retention period, but excludes the memory files in the me
+1 −1 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
interactive-mode
±changed
+2−1
claude-codechanged
interactive-mode
* Background tasks are automatically cleaned up when Claude Code exits. If you background the session instead of exiting it, Claude Code hands them to the background session, where
* Background tasks are automatically cleaned up when Claude Code exits. On macOS and Linux, when you stop a background task from [`/tasks`](/docs/en/commands) or Claude Code stops
* If you background the session instead of exiting it, your background tasks keep running in the background session. See [background a running session](/docs/en/agent-view#from-ins
+2 −1 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
headless
±changed
+7−3
claude-codechanged
headless
If Claude starts a [background Bash task](/docs/en/tools-reference#bash-tool-behavior) during a `claude -p` run, for example a dev server or a watch build, that shell is terminated
Background [subagents](/docs/en/sub-agents) and workflows are exempt from the five-second grace because their result is part of the final output, so `claude -p` waits for them to c
If Claude starts a [background Bash task](/docs/en/tools-reference#bash-tool-behavior) during a `claude -p` run, for example a dev server or a watch build, that shell is terminated
+7 −3 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
errors
±changed
+13−0
claude-codechanged
errors
| `Skill usage reports are not available on this connection.`
### Skill usage reports are not available on this connection
You ran [`/skill-doctor`](/docs/en/skills#find-unused-skills) over [Remote Control](/docs/en/remote-control), from your phone or browser. Claude Code doesn't send the skill usage r
+13 −0 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
env-vars
±changed
+4−3
claude-codechanged
env-vars
| `BASH_MAX_OUTPUT_LENGTH` | Maximum number of characters of bash output that Claude Code reads back into a command's result (default: 30000; maximum
| `BASH_MAX_OUTPUT_LENGTH` | Maximum number of characters of bash output that Claude Code reads back into a command's result (default: 30000; maximum
| `CLAUDE_CODE_PRINT_BG_WAIT_CEILING_MS` | Ceiling in milliseconds on continuous idle waiting after the final turn in [non-interactive mode](/docs/en/headless#back
| `CLAUDE_CODE_PRINT_BG_WAIT_CEILING_MS` | Ceiling in milliseconds on idle waiting for background subagents and workflows after the final turn in [non-interactive
| `TASK_MAX_OUTPUT_LENGTH` | Maximum number of characters in [subagent](/docs/en/sub-agents) output before truncation (default: 32000, maximum: 16000
+4 −3 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
discover-plugins
±changed
+2−1
claude-codechanged
discover-plugins
Run `/plugin` to open the plugin manager. This opens a tabbed interface with four tabs you can cycle through using **Tab**, or **Shift+Tab** to go backward:
Run `/plugin` to open the plugin manager. This opens a tabbed interface you can cycle through using **Tab**, or **Shift+Tab** to go backward:
* **Stats**: see [what each of your skills costs in context and how often it gets used](/docs/en/skills#find-unused-skills), in sessions where `/skill-doctor` is available
+2 −1 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
commands
±changed
+1−0
claude-codechanged
commands
| `/skill-doctor` | Show what each of your [skills](/docs/en/skills) costs in context and how often it
+1 −0 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
claude-directory
±changed
+31−24
claude-codechanged
claude-directory
| `debug/` | Per-session debug logs, written only when you st
| `debug/` | Per-session debug logs, written while debug logg
| `tasks/` | Per-session task lists written by the task tools
| `tasks/` | Task lists written by the task tools, one direct
* **Auto memory**: Claude Code excludes a project's [auto memory](/docs/en/memory#auto-memory) directory, `projects/<project>/memory/`, from this sweep, and removes the directory i
+31 −24 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
agent-sdk/user-input
±changed
+3−1
claude-codechanged
agent-sdk/user-input
Once you've passed a `canUseTool` callback in your query options, it fires when Claude wants to use a tool that nothing earlier in the permission flow has approved. Your callback r
Once you've passed a `canUseTool` callback in your query options, it fires when Claude wants to use a tool that nothing earlier in the permission flow has approved. In some configu
Your callback receives three arguments:
+3 −1 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
agent-sdk/typescript
±changed
+30−11
claude-codechanged
agent-sdk/typescript
| `permissionPrompts` | `'host' \| 'none'` | `'host'`
user_message_uuids?: string[];
Claude Code sets `user_message_uuid` on the turn's first assistant message, under the conditions in [`user_message_uuid`](#user_message_uuid).
Claude Code sets `user_message_uuid` and `user_message_uuids` on the turn's first assistant message, under the conditions in [`user_message_uuid`](#user_message_uuid).
user_message_uuids?: string[];
+30 −11 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
agent-sdk/permissions
±changed
+2−0
claude-codechanged
agent-sdk/permissions
In the TypeScript SDK, if you set [`permissionPrompts: 'none'`](/docs/en/agent-sdk/typescript#options), your callback isn't called at this step. A [`PermissionRequest` hook](/d
+2 −0 lines
Recorded here 5 Sep 2026 04:07 UTC
claude-code
agent-sdk/typescript
±changed
+10−2
claude-codechanged
agent-sdk/typescript
first_content_frame_ms?: number;
first_stream_post_ms?: number;
first_stream_post_ack_ms?: number;
first_stream_post_wall_ms?: number;
* `first_content_frame_ms`: time in milliseconds until the first `content_block_start` or `content_block_delta` stream event, counting thinking blocks as content. Present on the su
+10 −2 lines
Recorded here 5 Sep 2026 03:37 UTC
claude-code
errors
±changed
+26−4
claude-codechanged
errors
| `Restored the code, but skipped N files`
| `Restored the code, but skipped N files`
| `No files were restored: N files failed (backup missing, or the file could not be updated)`
## Rewind warnings
+26 −4 lines
Recorded here 5 Sep 2026 01:07 UTC
claude-code
checkpointing
±changed
+1−1
claude-codechanged
checkpointing
* Claude Code deletes checkpoints along with sessions after 30 days, following the [retention sweep rules](/docs/en/claude-directory#cleaned-up-automatically); change the period wi
* Claude Code deletes a session's file snapshots in the [retention sweep](/docs/en/claude-directory#cleaned-up-automatically), by default about 30 days after the session last saved
+1 −1 lines
Recorded here 5 Sep 2026 01:07 UTC
claude-code
troubleshooting
±changed
+14−0
claude-codechanged
troubleshooting
### Mouse wheel scrolls one line at a time in fullscreen rendering
In [fullscreen rendering](/docs/en/fullscreen), Claude Code scrolls the conversation itself rather than leaving it to your terminal. If each wheel notch moves fewer lines than you
To move faster without changing the speed, press `PgUp` and `PgDn` to scroll half a screen at a time. To hand scrolling back to your terminal's native scrollback instead, run `/tui
+14 −0 lines
Recorded here 5 Sep 2026 00:37 UTC
claude-code
sandboxing
±changed
+1−0
claude-codechanged
sandboxing
* **`pbcopy`, `xclip`, or `wl-copy` doesn't update the clipboard**: these clipboard utilities can fail to reach the system clipboard from inside the sandbox, in which case the text
+1 −0 lines
Recorded here 5 Sep 2026 00:37 UTC
claude-code
remote-control
±changed
+2−1
claude-codechanged
remote-control
A policy blocks Remote Control. The message's own text tells you which:
A policy blocks Remote Control. Check these causes in order:
* **Your claude.ai plan is Pro or Max**: Claude Code is still signed in under a Team or Enterprise organization from an earlier login, so it checks that organization's Remote Contr
+2 −1 lines
Recorded here 5 Sep 2026 00:37 UTC
4 Sep 2026
54 pages
+849 −748 lines
claude-code
statusline
±changed
+21−0
claude-codechanged
statusline
"last_miss_cause": {
"causes": ["tools_changed"],
"tools_added": 2,
"tools_removed": 0
},
+21 −0 lines
Recorded here 4 Sep 2026 23:37 UTC
claude-code
prompt-caching
±changed
+2−0
claude-codechanged
prompt-caching
The `Prompt cache (main)` line also names the likely cause of the last miss when Claude Code can identify one, for example `likely cause: tool definitions changed`. The likely-caus
+2 −0 lines
Recorded here 4 Sep 2026 23:37 UTC
claude-code
costs
±changed
+1−1
claude-codechanged
costs
* **Misses**: requests that re-processed content the cache already held, with the time of the last miss and how many tokens those requests wrote back to the cache. Claude Code coun
* **Misses**: requests that re-processed content the cache already held, with the time of the last miss and how many tokens those requests wrote back to the cache. Claude Code coun
+1 −1 lines
Recorded here 4 Sep 2026 23:37 UTC
claude-code
changelog
±changed
+70−0
claude-codechanged
changelog
<Update label="2.1.261" description="September 4, 2026">
* Added an "Organization policy" line to `/status` and `claude doctor` that says why your organization's policy could not be loaded, such as a proxy not passing the endpoint thro
* Added `bashOutputMaxChars` and `taskOutputMaxChars` settings to raise how much command and background-task output Claude receives inline before it is saved to a file, up to 128
* Added `--append-subagent-system-prompt-file` to read the subagent system prompt from a file, for prompts too large to pass on the command line
+70 −0 lines
Recorded here 4 Sep 2026 20:07 UTC
claude-code
keybindings
±changed
+21−1
claude-codechanged
keybindings
| `DiffPanel` | The [diff panel](/docs/en/interactive-mode#diff-panel) is open |
### Diff panel actions
Actions for the [diff panel](/docs/en/interactive-mode#diff-panel) that `/diff` opens in fullscreen rendering. `app:cycleDiffBase` is in the `DiffPanel` context, which is active wh
+21 −1 lines
Recorded here 4 Sep 2026 16:37 UTC
claude-code
interactive-mode
±changed
+43−0
claude-codechanged
interactive-mode
## Review changes with /diff
Run `/diff` to look over the changes in your working tree without leaving Claude Code. You see the edits Claude has made so far alongside anything else you haven't committed.
In [fullscreen rendering](/docs/en/fullscreen), `/diff` opens the [diff panel](#diff-panel) beside the conversation, which stays open and updates while you keep working. In the cla
+43 −0 lines
Recorded here 4 Sep 2026 16:37 UTC
claude-code
fullscreen
±changed
+4−0
claude-codechanged
fullscreen
## Watch your changes in the diff panel
In fullscreen rendering, [`/diff`](/docs/en/interactive-mode#review-changes-with-%2Fdiff) opens a panel beside the conversation rather than a viewer you have to close, so you can w
+4 −0 lines
Recorded here 4 Sep 2026 16:37 UTC
claude-code
commands
±changed
+1−1
claude-codechanged
commands
| `/diff` | Open an interactive diff viewer showing uncommitted changes and per-turn diffs. Us
| `/diff` | Review the changes in your working tree, including the edits Claude has made so fa
+1 −1 lines
Recorded here 4 Sep 2026 16:37 UTC
claude-code
agent-sdk/overview
±changed
+1−0
claude-codechanged
agent-sdk/overview
* [Migration guide](/docs/en/agent-sdk/migration-guide): migrate from the Claude Code SDK packages to the Agent SDK
+1 −0 lines
Recorded here 4 Sep 2026 15:37 UTC
claude-code
agent-sdk/migration-guide
±changed
+2−0
claude-codechanged
agent-sdk/migration-guide
Migrating from the OpenAI Agents SDK instead? The [OpenAI Agents SDK migration recipe](https://platform.claude.com/cookbook/claude-agent-sdk-04-migrating-from-openai-agents-sdk) ma
+2 −0 lines
Recorded here 4 Sep 2026 15:37 UTC
claude-code
agent-sdk/migration-guide
±changed
+5−5
claude-codechanged
agent-sdk/migration-guide
| Aspect | Old | New |
| :------------------------- | :-------------------------- | :----------------------------------------------------------------------- |
| **Package Name (TS/JS)** | `@anthropic-ai/claude-code` | `@anthropic-ai/claude-agent-sdk` |
| **Python Package** | `claude-code-sdk` | `claude-agent-sdk` |
| **Documentation Location** | Claude Code docs | Claude Code docs → dedicated [Agent SDK](/docs/en/agent-sdk/overview) section |
+5 −5 lines
Upstream edited 4 Sep 2026 15:09 UTC Recorded here 28 Sep 2026 23:37 UTC
claude-code
permission-modes
±changed
+9−0
claude-codechanged
permission-modes
Claude Code v2.1.257 and later also block these by default:
* Requesting credentials from the cloud instance-metadata endpoint, such as `169.254.169.254`, or explicitly authenticating a cloud, cluster, or registry call with the machine's ow
* Reaching a public host by a route other than a direct request, such as a tunnel, a reverse shell, or a resolver or proxy configuration rewritten to point outside
* Reading credentials that belong to the host rather than to your task, such as node certificates or the node's container-registry auth
+9 −0 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
model-config
±changed
+1−1
claude-codechanged
model-config
On the Anthropic API, the `/model` picker lists a Fable model only after the server reports it available for your organization. When you type `/model fable` or a Fable model ID, Cl
On the Anthropic API, the `/model` picker lists a Fable model only after the server reports it available for your organization. When you type `/model fable` or a Fable model ID, Cl
+1 −1 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
keybindings
±changed
+2−2
claude-codechanged
keybindings
| `chat:clearInput` | Ctrl+L | Force a full screen redraw, preserving input and conversation
| `chat:clearScreen` | Cmd+K | Force a full screen redraw, preserving input and conversation. See [Clear the conversation](/docs/en/fullscreen#clear-
| `chat:clearInput` | Ctrl+L | Force a full screen redraw, preserving input and conversation. In [fullscreen rendering](/docs/en/fullscreen#clear-the
| `chat:clearScreen` | Cmd+K | Same as `chat:clearInput`. See [Clear the conversation](/docs/en/fullscreen#clear-the-conversation) for how Cmd+K beha
+2 −2 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
interactive-mode
±changed
+1−1
claude-codechanged
interactive-mode
| `Ctrl+L` | Redraw screen
| `Ctrl+L` | Redraw or clear the screen
+1 −1 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
fullscreen
±changed
+5−3
claude-codechanged
fullscreen
Run `/clear` to start a new conversation. Pressing `Ctrl+L` or `Cmd+K` doesn't clear the conversation; Claude Code redraws the screen and keeps it. Before v2.1.238, Claude Code ran
On iTerm2 and Terminal.app, your terminal handles `Cmd+K` itself and clears its own screen without telling Claude Code. Claude Code detects the cleared screen and repaints the conv
Run `/clear` to start a new conversation.
+5 −3 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
errors
±changed
+18−1
claude-codechanged
errors
| `task output swap refused (tasks dir moved or linked)`
These errors come from Claude's built-in tools. Claude corrects most tool errors on its own; the first two below need a change from you, because they come from a subagent definitio
These errors come from Claude's built-in tools. Claude corrects most tool errors on its own. When one needs a change from you, that error's **What to do** list says what to change.
<h3 id="task-output-swap-refused">
Task output swap refused
+18 −1 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
auto-mode-config
±changed
+2−1
claude-codechanged
auto-mode-config
* **Context slots**: describe your organization, stack, and security posture so the classifier reads the other rules in your context. Unlike the other two kinds, context slots have
* **Context slots**: describe your organization, stack, and security posture so the classifier reads the other rules in your context. Each defaults to `None configured` or to the c
* **Host containment**: defaults to an ordinary developer machine or CI runner with open internet. If Claude Code runs in a container, VM, or pod with an egress allow-list or nei
+2 −1 lines
Recorded here 4 Sep 2026 15:07 UTC
claude-code
model-config
±changed
+9−0
claude-codechanged
model-config
#### A new session starts on a different model than you picked
When you pick a model with `/model` and your next session starts on something else, these are the usual causes:
* **You chose it for one session.** Pressing `s` in the picker, launching with `--model`, and running `/model` in non-interactive mode all apply to the current session and leave yo
+9 −0 lines
Recorded here 4 Sep 2026 14:37 UTC
claude-code
agent-sdk/troubleshooting
±changed
+1−1
claude-codechanged
agent-sdk/troubleshooting
# Troubleshooting
# Troubleshoot the Agent SDK
+1 −1 lines
Recorded here 4 Sep 2026 14:37 UTC
claude-code
agent-sdk/todo-tracking
±changed
+3−1
claude-codechanged
agent-sdk/todo-tracking
On TypeScript Agent SDK 0.3.233 and later, or Python Agent SDK 0.2.139 and later, the following tools aren't available on Opus 4.8, Sonnet 5, Fable 5, Mythos 5, or later versions
On TypeScript Agent SDK 0.3.233 and later, or Python Agent SDK 0.2.139 and later, the following restriction applies.
The following tools aren't available on Opus 4.8, Sonnet 5, Fable 5, Mythos 5, or later versions of those families unless you opt in:
+3 −1 lines
Recorded here 4 Sep 2026 14:37 UTC
claude-code
routines
±changed
+1−1
claude-codechanged
routines
* **Environment variables**: provide values Claude can use during each run. They're [visible to anyone who uses the environment](/docs/en/cloud-environments#what-carries-over-f
* **Environment variables**: provide values Claude can use during each run. They're [visible to anyone who uses the environment](/docs/en/cloud-environments#what-carries-over-f
+1 −1 lines
Recorded here 4 Sep 2026 05:07 UTC
claude-code
cloud-environments
±changed
+25−23
claude-codechanged
cloud-environments
Each [cloud session](/docs/en/claude-code-on-the-web) runs in a cloud environment. You can configure an environment to allow or deny [network access](#access-levels), [set environm
Each [cloud session](/docs/en/claude-code-on-the-web) runs in a cloud environment. You can configure an environment to allow or deny [network access](#access-levels), [set environm
Select **Add cloud environment**, or hover over an existing environment and select the settings icon that appears on the right. The dialog includes the name, network access lev
Select **Add cloud environment**, or hover over an existing environment and select the settings icon that appears on the right. The dialog includes the name, network access lev
Anyone who uses the environment can read the values. For a key the agent proxy can attach to a request, use an [API credential](#add-api-credentials) instead. The [requests that ne
+25 −23 lines
Recorded here 4 Sep 2026 05:07 UTC
claude-code
claude-code-on-the-web
±changed
+1−1
claude-codechanged
claude-code-on-the-web
* **API credentials**: in Anthropic-hosted environments, keys you [add to a cloud environment](/docs/en/cloud-environments#add-api-credentials) stay outside the sandbox the same wa
* **API credentials**: in Anthropic-hosted environments on Pro and Max plans, keys you [add to a cloud environment](/docs/en/cloud-environments#add-api-credentials) stay outside th
+1 −1 lines
Recorded here 4 Sep 2026 05:07 UTC
claude-code
server-managed-settings
±changed
+2−1
claude-codechanged
server-managed-settings
* **Managed CLAUDE.md content**: a `claudeMd` value delivered through managed settings
A managed CLAUDE.md delivered through the [`claudeMd`](/docs/en/settings-reference#claudemd) key doesn't require approval, because it's instruction text for Claude rather than a co
+2 −1 lines
Recorded here 4 Sep 2026 03:37 UTC
claude-code
claude-apps-gateway-config
±changed
+0−1
claude-codechanged
claude-apps-gateway-config
* managed CLAUDE.md content
+0 −1 lines
Recorded here 4 Sep 2026 03:37 UTC
claude-code
prompt-caching
±changed
+7−3
claude-codechanged
prompt-caching
Two settings aren't part of the prompt text at all, so they don't appear in the layer table, but both are part of the cache key:
Two settings don't appear in the layer table but still affect what stays cached:
* **Effort level**: each effort level has its own cache for the same model. Changing effort mid-session recomputes the entire request. See [Changing effort level](#changing-effort-
* **Effort level**: on most models, each effort level has its own cache, so changing effort mid-session recomputes the entire request. On Fable 5.1 with an API key or a Claude subs
The cache is keyed by [effort level](/docs/en/model-config#adjust-effort-level) as well as model, so switching with `/effort` means the next request reads the entire conversation h
+7 −3 lines
Recorded here 4 Sep 2026 02:37 UTC
claude-code
llm-gateway-rollout
±changed
+3−3
claude-codechanged
llm-gateway-rollout
* For what Claude Code sends to a gateway and what to forward, see the [gateway protocol reference](/docs/en/llm-gateway-protocol)
* For what Claude Code sends to a gateway and what to forward, see the [gateway compatibility guide](/docs/en/llm-gateway-protocol)
When sizing per-key rate limits, account for the client [retrying transient failures](/docs/en/errors#automatic-retries), including `429` responses, up to 10 times with backoff, ho
When sizing per-key rate limits, account for the client [retrying transient failures](/docs/en/errors#automatic-retries), including `429` responses, up to 10 times with backoff, ho
* [Gateway protocol reference](/docs/en/llm-gateway-protocol): the wire contract for gateway operators, covering endpoints, headers to forward, and the feature pass-through table
+3 −3 lines
Recorded here 4 Sep 2026 02:37 UTC
Nothing on this page matches that. Press Enter to search every page of the ledger.