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
587 changed this week · last read 22 Oct 2026
Claude Code CLI
228pages tracked
146 changed this week · last read 22 Oct 2026
Claude Documentation
278pages tracked
74 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
5,287 of the ledger's changes in claude-code. Show every change
ReadSourcePageWhatLines
21 Oct 2026
72 pages
+1136 −321 lines
claude-code
agent-view
±changed
+31−5
claude-codechanged
agent-view
* **Foreground subagents are still running**: Claude Code keeps waiting so the work of the [foreground subagents](/docs/en/sub-agents#run-subagents-in-foreground-or-background) Cla
* **Foreground subagents are still running**: Claude Code keeps waiting so the work of the [foreground subagents](/docs/en/sub-agents#run-subagents-in-foreground-or-background) tha
* **You type into the prompt input**: Claude Code cancels the switch, because unsent text stays in your terminal's input box and wouldn't move to the background session. It shows `
* **You type into the prompt input**: Claude Code cancels the switch, because unsent text stays in your terminal's input box and wouldn't move to the background session. It shows `
+31 −5 lines
Upstream edited 21 Oct 2026 14:49 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
agent-sdk/subagents
±changed
+2−2
claude-codechanged
agent-sdk/subagents
| `skills` | `string[]` | No | List of skill names to preload into the agent's context at startup. Unlisted skills remain invocable through the Skill tool |
| `skills` | `string[]` | No | List of skill names to preload into the agent's context at startup. At most the [first 32 distinct names](/docs/en/sub-agents#preload-skills-into-sub
Messages from within a subagent's context include a `parent_tool_use_id` field. In TypeScript, each assistant and user message a subagent produces also carries [`agent_id`](/docs/e
Messages from within a subagent's context include a `parent_tool_use_id` field. In TypeScript, each assistant and user message a subagent produces also carries [`agent_id`](/docs/e
+2 −2 lines
Upstream edited 21 Oct 2026 14:49 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
managed-settings
±changed
+1−1
claude-codechanged
managed-settings
Wherever the session runs, claude.ai applies the admin console's [`strictKnownMarketplaces`](/docs/en/settings-reference#strictknownmarketplaces) and [`blockedMarketplaces`](/doc
Wherever the session runs, claude.ai applies the admin console's [`strictKnownMarketplaces`](/docs/en/settings-reference#strictknownmarketplaces) and [`blockedMarketplaces`](/doc
+1 −1 lines
Upstream edited 21 Oct 2026 14:48 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
desktop-quickstart
±changed
+10−10
claude-codechanged
desktop-quickstart
The desktop app gives you Claude Code with a graphical interface, so you can ask Claude to work on the code in a folder on your computer and review its changes without using a term
The desktop app gives you Claude Code with a graphical interface. You can ask Claude to work on the code in a folder on your computer and review its changes without using a termina
Claude Code is in the **Code** tab at the top of the app's sidebar. This page walks through installing the app and starting your first session there. Claude Code requires a [Pro, M
+10 −10 lines
Upstream edited 21 Oct 2026 14:48 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
desktop-linux
±changed
+5−5
claude-codechanged
desktop-linux
Cowork is the desktop tab for [Dispatch and longer agentic work](https://claude.com/docs/cowork/overview). On Linux, Cowork runs those tasks in a virtual machine that the desktop a
Cowork is the part of the desktop app for [Dispatch and longer agentic work](https://claude.com/docs/cowork/overview). On Linux, Cowork runs those tasks in a virtual machine that t
* **Hardware virtualization**: turned on in your firmware settings. Without it, the Cowork tab reports "Cowork requires hardware virtualization (KVM)".
* **QEMU and UEFI firmware**: `qemu-system-x86`, `ovmf`, and `virtiofsd` on x86\_64, or `qemu-system-arm`, `qemu-efi-aarch64`, and `virtiofsd` on arm64. `apt install claude-desktop
* **Hardware virtualization**: turned on in your firmware settings. Without it, Cowork reports "Cowork requires hardware virtualization (KVM)".
+5 −5 lines
Upstream edited 21 Oct 2026 14:48 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
desktop
±changed
+13−6
claude-codechanged
desktop
The Claude Desktop app has three tabs: **Chat** for conversations, **Cowork** for [Dispatch and longer agentic work](https://claude.com/product/cowork), and **Code** for software d
In the Claude Desktop app, Claude Code is the **Code** tab. The app also has chat and [Cowork](https://claude.com/product/cowork), which handles Dispatch and longer agentic work. T
The session details panel is a set of cards beside the chat that show where the session runs, what it changed, and what it's running. The panel appears after you start a session, w
The session details panel is a set of cards beside the chat that show where the session runs, what it changed, what it's running, and its context usage. The panel appears after you
Click the usage ring next to the model picker to see your current context window usage and your plan usage for the period. Context usage is per session; plan usage is shared across
+13 −6 lines
Upstream edited 21 Oct 2026 14:48 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
server-managed-settings
±changed
+1−1
claude-codechanged
server-managed-settings
In a [Cowork](https://claude.com/docs/cowork/overview) session in the Claude Desktop app, Claude Code doesn't fetch server-managed settings from the claude.ai admin console, even w
In a [Cowork](https://claude.com/docs/cowork/overview) session in the Claude Desktop app, Claude Code doesn't fetch server-managed settings from the claude.ai admin console, even w
+1 −1 lines
Upstream edited 21 Oct 2026 14:47 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
permission-modes
±changed
+1−1
claude-codechanged
permission-modes
The Cowork tab doesn't use these modes. Cowork has its own permission modes, enabled separately, and the Cowork tab shows no mode selector at all until a mode beyond its default is
Cowork doesn't use these modes. It has its own permission modes, enabled separately, and shows no mode selector at all until a mode beyond its default is enabled for your account.
+1 −1 lines
Upstream edited 21 Oct 2026 14:47 UTC Recorded here 21 Oct 2026 15:07 UTC
claude-code
claude-tag
±changed
+2−2
claude-codechanged
claude-tag
No line of the body moved in this read.
+2 −2 lines
Recorded here 21 Oct 2026 10:07 UTC
claude-code
skills
±changed
+5−5
claude-codechanged
skills
| Project | `.claude/skills/<skill-name>/SKILL.md` | Sessions in this repository. Commit it so your team gets it too |
| Nested | `<subdir>/.claude/skills/<skill-name>/SKILL.md` | Sessions started in or below `<subdir>`. A session started above it loads the skill once Claude works on files there. S
| Project | `.claude/skills/<skill-name>/SKILL.md` | Sessions in this repository, but not [cloud threads in a project](/docs/en/claude-projects#what-threads-pick-up-from-your-repos
| Nested | `<subdir>/.claude/skills/<skill-name>/SKILL.md` | Sessions started in or below `<subdir>`, but not cloud threads in a project. Outside those threads, a session started a
[Cowork](https://claude.com/product/cowork) sessions and [cloud sessions](/docs/en/cloud-environments#what-carries-over-from-your-setup), including [cloud scheduled tasks](/docs/en
+5 −5 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
settings-example
±changed
+1−1
claude-codechanged
settings-example
* **Cloud sessions read it too.** A [cloud session](/docs/en/settings#settings-in-cloud-sessions) starts from a clone of the repository, so the committed file applies there as well
* **Cloud sessions outside a [project](/docs/en/claude-projects#what-threads-pick-up-from-your-repositories) read it too.** A [cloud session](/docs/en/settings#settings-in-cloud-se
+1 −1 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
self-hosted-environments-configuration
±changed
+7−5
claude-codechanged
self-hosted-environments-configuration
* `<repo>/.mcp.json`: project scope. Commit the file to the repository; its servers are auto-approved in cloud sessions. In a session with several repositories, [at most one reposi
* `<repo>/.mcp.json`: project scope. Outside a [project](/docs/en/claude-projects#what-threads-pick-up-from-your-repositories) thread, commit the file to the repository, and its se
* **Project settings**: a repository-committed `.claude/settings.json` layers on top of the user-level baseline. In a session with several repositories, [at most one repository's f
* **Project settings**: outside a [project](/docs/en/claude-projects#what-threads-pick-up-from-your-repositories) thread, a repository-committed `.claude/settings.json` layers on t
When a person starts their own session, Claude Code also downloads the [skills enabled for their claude.ai account](/docs/en/skills#skills-in-cowork-and-cloud-sessions) into that s
+7 −5 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
routines
±changed
+2−2
claude-codechanged
routines
Scheduled tasks run autonomously as full Claude Code cloud sessions. You don't choose a permission mode. The session runs shell commands, uses [skills](/docs/en/skills) committed t
Scheduled tasks run autonomously as full Claude Code cloud sessions. You don't choose a permission mode. The session runs shell commands, uses [skills](/docs/en/skills), and calls
* **In the repository**: for a scheduled task with one repository, you can instead declare the server in a committed [`.mcp.json`](/docs/en/mcp#project-scope) so it is part of the
* **In the repository**: for a scheduled task with one repository outside a [project](/docs/en/claude-projects#what-threads-pick-up-from-your-repositories), you can instead declare
+2 −2 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
mcp-quickstart
±changed
+1−1
claude-codechanged
mcp-quickstart
* **Cloud sessions**: commit a `.mcp.json` to your repository; a session with one repository loads it. See [Edit .mcp.json directly](#edit-mcp-json-directly) and [What carries over
* **Cloud sessions**: outside a [project](/docs/en/claude-projects#what-threads-pick-up-from-your-repositories), commit a `.mcp.json` to your repository; a session with one reposit
+1 −1 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
mcp
±changed
+1−1
claude-codechanged
mcp
In `claude -p` runs, [Agent SDK](/docs/en/headless) sessions, and [cloud sessions](/docs/en/claude-code-on-the-web), Claude Code can't show that prompt: it loads project-scoped ser
In `claude -p` runs, [Agent SDK](/docs/en/headless) sessions, and [cloud sessions](/docs/en/claude-code-on-the-web) outside a [project](/docs/en/claude-projects#what-threads-pick-u
+1 −1 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
cloud-environments
±changed
+7−2
claude-codechanged
cloud-environments
> [!NOTE]
>
> Of the repository configuration in this table, a cloud thread in a [project](/docs/en/claude-projects) loads only the repository root's `CLAUDE.md`, `.claude/CLAUDE.md`, and `.cl
* **Anthropic-hosted environment**: Claude Code runs hooks from the repository and from your organization's [server-managed settings](/docs/en/server-managed-settings). [Claude Tag
+7 −2 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
claude-projects
±changed
+20−24
claude-codechanged
claude-projects
* The `CLAUDE.md` and skills in [each of the project's repositories](#what-threads-pick-up-from-your-repositories), and in a project with one repository, that repository's permis
* The `CLAUDE.md` in [each of the project's repositories](#what-threads-pick-up-from-your-repositories), but not the repositories' skills, `.claude/settings.json`, or `.mcp.json`
Each approval covers that prompt, or the rest of that thread if you choose the broader option.
To let every thread run certain commands without asking, or to block some, add [permission rules](/docs/en/permissions) to the repository's `.claude/settings.json`. Check that clou
+20 −24 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
claude-code-on-the-web
±changed
+2−2
claude-codechanged
claude-code-on-the-web
* **`/config`**: in your browser at claude.ai/code, opens the Claude Code section of your settings instead of setting a value, and text after the command, including `key=value`, is
* **`/config`**: in your browser at claude.ai/code, opens the Claude Code section of your settings instead of setting a value, and text after the command, including `key=value`, is
[Subagents](/docs/en/sub-agents) work the same way they do locally. Claude can spawn them with the Agent tool to offload research or parallel work into a separate context window, k
[Subagents](/docs/en/sub-agents) work the same way they do locally. Claude can spawn them with the Agent tool to offload research or parallel work into a separate context window, k
+2 −2 lines
Upstream edited 21 Oct 2026 10:03 UTC Recorded here 21 Oct 2026 10:07 UTC
claude-code
claude-tag
±changed
+2−2
claude-codechanged
claude-tag
No line of the body moved in this read.
+2 −2 lines
Recorded here 21 Oct 2026 06:37 UTC
claude-code
agent-sdk/mcp
±changed
+6−0
claude-codechanged
agent-sdk/mcp
### Named servers
### Set the first-turn wait yourself
+6 −0 lines
Upstream edited 21 Oct 2026 06:28 UTC Recorded here 21 Oct 2026 06:37 UTC
claude-code
setup
±changed
+5−5
claude-codechanged
setup
```bash theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} the
```bash theme={null}
```powershell theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={nul
```powershell theme={null}
```batch theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} th
+5 −5 lines
Upstream edited 21 Oct 2026 06:28 UTC Recorded here 21 Oct 2026 06:37 UTC
claude-code
prompt-caching
±changed
+1−1
claude-codechanged
prompt-caching
* **Tools loaded upfront**: adding a definition invalidates the cache, and so does removing one on purpose. This applies when tool search is [below its `auto` threshold, disabled,
* **Tools loaded upfront**: adding a definition invalidates the cache, and so does removing one on purpose. This applies when tool search is [off or unavailable](/docs/en/mcp#confi
+1 −1 lines
Upstream edited 21 Oct 2026 06:28 UTC Recorded here 21 Oct 2026 06:37 UTC
claude-code
mcp
±changed
+5−1
claude-codechanged
mcp
| `auto` | Threshold mode: Claude Code loads the tools it would otherwise defer upfront while their definitions total less than 10% of the context window, and defers all of them on
| `auto` | Threshold mode. Tools that would otherwise be deferred load upfront while their definitions total under 10% of the context window. At 10% or more, tool search turns on,
#### Tools that stay deferred below the `ENABLE_TOOL_SEARCH=auto` threshold
+5 −1 lines
Upstream edited 21 Oct 2026 06:28 UTC Recorded here 21 Oct 2026 06:37 UTC
claude-code
agent-sdk/tool-search
±changed
+1−1
claude-codechanged
agent-sdk/tool-search
| `auto` | Counts the tokens in the tool definitions that tool search can defer and compares the total against the model's context window. When the total reaches 10% of the window,
| `auto` | Tool search is off, and tool definitions load upfront, until the definitions it can defer reach 10% of the model's context window. Tool search then activates, and can [s
+1 −1 lines
Upstream edited 21 Oct 2026 06:28 UTC Recorded here 21 Oct 2026 06:37 UTC
claude-code
sandboxing
±changed
+13−3
claude-codechanged
sandboxing
If the reason in parentheses is `EACCES`, `EPERM`, or `EROFS`, start Claude Code in a project folder you can write to, then run `/sandbox` again in that session.
If the reason begins with `Invalid JSON syntax in settings file`, correct the JSON in your project's `.claude/settings.local.json`, then run `/sandbox` again.
If the reason in parentheses is `EACCES`, `EPERM`, or `EROFS`, start Claude Code in a project folder you can write to, then run `/sandbox` again in that session. If the reason begi
+13 −3 lines
Upstream edited 21 Oct 2026 05:51 UTC Recorded here 21 Oct 2026 06:07 UTC
claude-code
permissions
±changed
+26−3
claude-codechanged
permissions
* **`find`**: with an action that runs commands, deletes files, or writes files, such as `-exec`, `-delete`, or `-fprint`, or with `-files0-from`, which takes the paths to search f
* **`find`**: with an action that runs commands, deletes files, or writes files, such as `-exec`, `-delete`, `-rm`, or `-fprint`, or with `-files0-from`, which takes the paths to s
##### Attachments through a symlink
A `Read` deny rule also covers these attachments, matching each one on either the named path or the file it resolves to. Claude Code leaves an attachment out of the conversation wh
+26 −3 lines
Upstream edited 21 Oct 2026 05:51 UTC Recorded here 21 Oct 2026 06:07 UTC
claude-code
permission-modes
±changed
+1−1
claude-codechanged
permission-modes
> In the VS Code extension, how [Claude in Chrome](/docs/en/chrome) browser actions get approved depends on how the session connected to the browser: see [Permission prompts in VS
> For how [Claude in Chrome](/docs/en/chrome) browser actions get approved in the VS Code extension, see [Permission prompts in VS Code sessions](/docs/en/chrome#permission-prompts
+1 −1 lines
Upstream edited 21 Oct 2026 05:51 UTC Recorded here 21 Oct 2026 06:07 UTC
claude-code
errors
±changed
+24−1
claude-codechanged
errors
| `Only a teammate can approve a shutdown request` | [Tool errors](#only-a-teammate-can-approve-a-shutdown-request) |
| `CLAUDE_CODE_AUTO_COMPACT_WINDOW="<value>" reads as <N> tokens, so Claude Code treats it as the 100k minimum` | [Values and warnings for `CLAUDE_CODE_AUTO_COMPACT_WINDOW`](/docs/
| `CLAUDE_CODE_AUTO_COMPACT_WINDOW="<value>" is too large, so Claude Code treats it as the 1M maximum` | [Values and warnings for `CLAUDE_CODE_AUTO_COMPACT_WINDOW`](/docs/en/model-
| `CLAUDE_CODE_AUTO_COMPACT_WINDOW="<value>" isn't a usable size, so Claude Code ignores it` | [Values and warnings for `CLAUDE_CODE_AUTO_COMPACT_WINDOW`](/docs/en/model-config#cla
* If [`ANTHROPIC_BASE_URL`](/docs/en/env-vars) points at an [LLM gateway](/docs/en/llm-gateway) and every request fails this way, check whether your gateway is rejecting the tool s
+24 −1 lines
Upstream edited 21 Oct 2026 05:25 UTC Recorded here 21 Oct 2026 05:37 UTC
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 21 Oct 2026 03:37 UTC
claude-code
changelog
±changed
+5−0
claude-codechanged
changelog
**2.1.306**
* Fixed `claude -p` inside cloud sessions reporting "Not logged in" when started without the session's environment variables
* Fixed an API error being replaced by "prompt is too long", and an unneeded compaction, in sessions with many MCP tools
+5 −0 lines
Upstream edited 21 Oct 2026 03:27 UTC Recorded here 21 Oct 2026 03:37 UTC
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 21 Oct 2026 02:37 UTC
claude-code
code-review-troubleshooting
±changed
+5−1
claude-codechanged
code-review-troubleshooting
Two of the messages a check run can carry about `REVIEW.md` concern branch protection:
These are three of the messages a check run can carry about `REVIEW.md`:
* **"they link to a file this pull request modifies, or carry more links than the review checks"**: the review skipped the root `REVIEW.md`, or the number of subdirectory `REVIEW.m
* **Too many links**: it carries more than 64 Markdown links. Cut links to [64 or fewer](/docs/en/code-review-customize#include-other-files-in-review-md).
* **A link the review can't parse**: it has a malformed Markdown link. Fix or remove the link.
+5 −1 lines
Upstream edited 21 Oct 2026 02:12 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
code-review-customize
±changed
+40−2
claude-codechanged
code-review-customize
### Include other files in REVIEW\.md
You can split review instructions across several files and pull them into your root `REVIEW.md`. Use this when `REVIEW.md` grows long or when you already keep review rules in anoth
To include a file, add a line to the root `REVIEW.md` with only `@` and the file name on it. Code Review puts the file's contents in place of that line, so they apply as part of `R
+40 −2 lines
Upstream edited 21 Oct 2026 02:12 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
settings-reference
±changed
+2−2
claude-codechanged
settings-reference
See summaries of Claude's [extended thinking](/docs/en/model-config#extended-thinking) in interactive sessions. Set it if you want the full summaries when you expand thinking with
See summaries of Claude's [extended thinking](/docs/en/model-config#extended-thinking) in interactive sessions. Set it if you want the full summaries when you expand thinking with
Choose how Claude Code notifies you when a task completes or a permission prompt is waiting. Appears in `/config` as **Local notifications**.
Choose how Claude Code notifies you when a task completes or a permission prompt is waiting. A [mod's notifications](/docs/en/plugins/mods/api#show-something-without-starting-a-tur
+2 −2 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
plugins/mods/test
±changed
+1−0
claude-codechanged
plugins/mods/test
| `$.ui.notify` | `() => ({ value: { isSent: true, channel: 'terminal_bell' } })`. The text is `e.text` and the title is `e.title`. |
+1 −0 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
plugins/mods/reference
±changed
+1−1
claude-codechanged
plugins/mods/reference
| [`$.ui`](/docs/en/plugins/mods/interface#pick-where-to-draw) | `resolve`, `invalidate`, `open`, `close`, `panes`, `focus`, `scroll`, `toast`, `status`, `log`, `notice`, `ask`, `c
| [`$.ui`](/docs/en/plugins/mods/interface#pick-where-to-draw) | `resolve`, `invalidate`, `open`, `close`, `panes`, `focus`, `scroll`, `toast`, `notify`, `status`, `log`, `notice`,
+1 −1 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
plugins/mods/api
±changed
+36−0
claude-codechanged
plugins/mods/api
| `$.ui.notify(text)` | A notification from their terminal, outside Claude Code, sent the way their [`preferredNotifChannel`](/docs/en/settings-reference#preferrednotifchannel) set
`$.ui.notify` is the only one of these that answers. It resolves to `{ isSent: true, channel }` when the user's channel wrote the notification to their terminal. `channel` is the s
Otherwise the call resolves to `{ isSent: false, reason }`:
+36 −0 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
hooks-guide
±changed
+2−1
claude-codechanged
hooks-guide
| `plugin_notification` | A [mod](/docs/en/plugins/mods/api#show-something-without-starting-a-turn) raises a notification with `$.ui.notify`. The hook's `title` is the mod's name u
| `Notification` | notification type | `permission_prompt`, `idle_prompt`, `auth_success`, `elicitation_dialog`, `elicitation_url_dialog`, `elicitation_complete`, `elicitation_resp
| `Notification` | notification type | `permission_prompt`, `idle_prompt`, `auth_success`, `elicitation_dialog`, `elicitation_url_dialog`, `elicitation_complete`, `elicitation_resp
+2 −1 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
hooks
±changed
+2−1
claude-codechanged
hooks
| `Notification` | notification type | `permission_prompt`, `idle_prompt`, `auth_success`, `elicitation_dialog`, `elicitation_url_dialog`, `elicitation_complete`, `elicitation_resp
| `Notification` | notification type | `permission_prompt`, `idle_prompt`, `auth_success`, `elicitation_dialog`, `elicitation_url_dialog`, `elicitation_complete`, `elicitation_resp
| `plugin_notification` | A [mod](/docs/en/plugins/mods/api#show-something-without-starting-a-turn) raises a notification with `$.ui.notify`. The hook's `title` is the mod's name u
+2 −1 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
agent-sdk/hooks
±changed
+1−0
claude-codechanged
agent-sdk/hooks
* `plugin_notification` when a [mod](/docs/en/plugins/mods/api#show-something-without-starting-a-turn) calls `$.ui.notify`
+1 −0 lines
Upstream edited 21 Oct 2026 02:10 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
prompt-caching
±changed
+1−1
claude-codechanged
prompt-caching
* **Rejects the marked request with a `400` error naming `cache_control`**: Claude Code re-sends the request with the marker moved off the block and onto your last conversation mes
* **Rejects the marked request with a `400` error naming `cache_control`**: Claude Code re-sends the request with the marker moved off the block and onto your last conversation mes
+1 −1 lines
Upstream edited 21 Oct 2026 02:07 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
llm-gateway-protocol
±changed
+5−3
claude-codechanged
llm-gateway-protocol
When the upstream rejects one of these capabilities, Claude Code can retry the request on its own without the rejected part:
* When the upstream rejects the `thinking` field, a mid-conversation system message, or the `cache_control` marker on such a message, Claude Code retries the request and disables t
When the upstream rejects one of these capabilities, Claude Code can retry the request on its own with the rejected part removed or changed:
+5 −3 lines
Upstream edited 21 Oct 2026 02:07 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
cross-session-messaging
±changed
+1−1
claude-codechanged
cross-session-messaging
Set `dialogExpiry` to `"never"` to keep default-held messages until the session ends. A message held by an explicit `hold` setting doesn't expire; Claude Code delivers it only when
Set `dialogExpiry` to `"never"` to keep default-held messages until the session ends. A message held by an explicit `hold` setting doesn't expire; Claude Code delivers it when an `
+1 −1 lines
Upstream edited 21 Oct 2026 02:07 UTC Recorded here 21 Oct 2026 02:37 UTC
claude-code
claude-tag
±changed
+2−2
claude-codechanged
claude-tag
No line of the body moved in this read.
+2 −2 lines
Recorded here 21 Oct 2026 02:07 UTC
claude-code
hooks-guide
±changed
+1−1
claude-codechanged
hooks-guide
A fourth value, `"defer"`, is available in [non-interactive mode](/docs/en/headless) with the `-p` flag. It exits the process with the tool call preserved so an Agent SDK wrapper c
A fourth value, `"defer"`, is available in [non-interactive mode](/docs/en/headless) with the `-p` flag. Returning it ends the turn with the tool call preserved, so an Agent SDK wr
+1 −1 lines
Upstream edited 21 Oct 2026 02:01 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
hooks
±changed
+10−5
claude-codechanged
hooks
| `permissionDecision` | `"allow"` skips the permission prompt, except for [actions no mode auto-approves](/docs/en/permission-modes#actions-no-mode-auto-approves) and [reads from
| `permissionDecision` | `"allow"` skips the permission prompt, except for [actions no mode auto-approves](/docs/en/permission-modes#actions-no-mode-auto-approves) and [reads from
2. The hook returns `permissionDecision: "defer"`. The tool doesn't execute. The process exits with `stop_reason: "tool_deferred"` and the pending tool call preserved in the transc
2. The hook returns `permissionDecision: "defer"`. The tool doesn't execute. The turn ends with a result carrying `stop_reason: "tool_deferred"`, and the pending tool call stays in
There is no timeout or retry limit. The session remains on disk until you resume it, subject to the [`cleanupPeriodDays`](/docs/en/settings-reference#cleanupperioddays) retention s
+10 −5 lines
Upstream edited 21 Oct 2026 02:01 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
agent-sdk/typescript
±changed
+2−0
claude-codechanged
agent-sdk/typescript
The `StructuredOutput` tool has no type here: Claude Code builds its input schema from the schema you pass to `outputFormat`, so there is nothing fixed to export. See [Recognize th
+2 −0 lines
Upstream edited 21 Oct 2026 02:01 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
agent-sdk/structured-outputs
±changed
+11−0
claude-codechanged
agent-sdk/structured-outputs
## Recognize the StructuredOutput tool call
When you set `outputFormat` in TypeScript or `output_format` in Python, your message stream carries a tool call named `StructuredOutput` that you never defined. Claude Code adds th
* A `tool_use` block named `StructuredOutput`, whose input is the object the agent produced for your schema
+11 −0 lines
Upstream edited 21 Oct 2026 02:01 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
agent-sdk/streaming-output
±changed
+1−1
claude-codechanged
agent-sdk/streaming-output
* **Structured output**: with partial messages enabled, the JSON streams as a tool call's unvalidated `input_json_delta` chunks, and only the validated result reaches the final `Re
* **Structured output**: with partial messages enabled, the JSON streams as a tool call's unvalidated `input_json_delta` chunks, and only the validated result reaches the final `Re
+1 −1 lines
Upstream edited 21 Oct 2026 02:01 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
permission-modes
±changed
+20−5
claude-codechanged
permission-modes
<video autoPlay muted loop playsInline src="https://mintcdn.com/claude-code/oa7CKjMeIChox26S/images/permission-modes-cycle-light.mp4?fit=max&auto=format&n=oa7CKjMeIChox26S&q=85&s=1
<video autoPlay muted loop playsInline src="https://mintcdn.com/claude-code/tA-x-5L2YsYSVdee/images/permission-modes-cycle-light.mp4?fit=max&auto=format&n=tA-x-5L2YsYSVdee&q=85&s=3
In [non-interactive mode](/docs/en/headless) no dialog is shown. A [background session](/docs/en/agent-view) honors your acceptance when it's recorded in user or managed settings:
* With no acceptance recorded, `claude --bg --permission-mode bypassPermissions` is refused until you accept the dialog in an interactive session.
+20 −5 lines
Upstream edited 21 Oct 2026 01:41 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
errors
±changed
+1−1
claude-codechanged
errors
| `Bypass permissions was requested at launch and ignored` | [Skip all checks with bypassPermissions mode](/docs/en/permission-modes#bypass-permissions-acceptance) |
| `Bypass permissions was requested at launch and ignored` | [Bypass permissions ignored in a background session](/docs/en/permission-modes#bypass-permissions-ignored-in-a-backgrou
+1 −1 lines
Upstream edited 21 Oct 2026 01:41 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
agent-view
±changed
+5−8
claude-codechanged
agent-view
The [permission mode](/docs/en/permissions) depends on how you started the session:
* **Backgrounded with `/bg` or `←`**: Claude Code keeps the permission mode the session was in, so one you switched to `acceptEdits` or `auto` stays there after detaching
* **Dispatched from an agent view you opened with `←`**: Claude Code takes the new session's permission mode from the target directory's configuration first, then from the session
* **Backgrounded with `/bg` or `←`**: Claude Code keeps the [permission mode](/docs/en/permissions) the session was in, so one you switched to `acceptEdits` or `auto` stays there a
+5 −8 lines
Upstream edited 21 Oct 2026 01:41 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
agent-sdk/python
±changed
+13−6
claude-codechanged
agent-sdk/python
* `is_error`: `True` when the conversation ended in an error state. Always `True` on the `error_*` subtypes. On `subtype="success"` it is `True` when the final model request failed
* `is_error`: `True` when the conversation ended in an error state. Always `True` on the `error_*` subtypes. On `subtype="success"` it is `True` in three cases. Check `stop_reason`
* `"refusal"`: the model declined to respond. Read the refusal text in `result`, or in the preceding `AssistantMessage` when `result` is empty.
* `"tool_deferred_unavailable"`: the [deferred tool](/docs/en/hooks#defer-a-tool-call-for-later) was no longer available when you resumed the session, for example because its MCP
* No prompt: the call is still pending. Make the tool available and resume again without a prompt, and the call runs first.
+13 −6 lines
Upstream edited 21 Oct 2026 01:41 UTC Recorded here 21 Oct 2026 02:07 UTC
claude-code
claude-tag
±changed
+2−2
claude-codechanged
claude-tag
No line of the body moved in this read.
+2 −2 lines
Recorded here 21 Oct 2026 01:37 UTC
claude-code
self-hosted-environments-configuration
±changed
+16−2
claude-codechanged
self-hosted-environments-configuration
| `CLAUDE_RUNNER_ACCOUNT_ID` | Tagged ID of the account that enqueued the session, for per-account routing, quota, or chargeback. Empty when unavailable, and always empty for Claud
| `CLAUDE_RUNNER_ACCOUNT_ID` | Tagged ID of the account that enqueued the session. Use it for per-account routing, quota, or chargeback, not as a dedup key. Empty when unavailable,
* **Commands that are allowed to fail**: if the hook didn't use `set -e` before, it now stops at the first command that returns non-zero, such as a lookup that finds nothing or a d
* **Commands that are allowed to fail**: if the hook didn't use `set -e` before, it now stops at the first command that returns non-zero, such as a lookup that finds nothing. If th
* **The command that submits the workload**: don't follow it with `|| true`. With `|| true`, the hook exits 0 even when your platform refuses the submit, for example over quota, an
+16 −2 lines
Upstream edited 21 Oct 2026 01:36 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
permissions
±changed
+3−3
claude-codechanged
permissions
Allow rules accept tool-name globs only after a literal `mcp__<server>__` prefix. The server segment must be glob-free so the rule names a specific server you configured. `mcp__pup
A deny or ask rule whose tool name matches no known tool produces a startup warning to catch typos. Tool names containing `_` or `*` are exempt from the check, and so are the names
Allow rules accept tool-name globs only after a literal `mcp__<server>__` prefix. The server segment must be glob-free so the rule names a specific server you configured. `mcp__pup
+3 −3 lines
Upstream edited 21 Oct 2026 01:33 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
model-config
±changed
+6−2
claude-codechanged
model-config
Organization admins on Claude Enterprise plans can set a maximum [effort level](#adjust-effort-level) per model for each custom role, alongside role-level [organization model restr
Organization admins on Claude Enterprise plans can set a maximum [effort level](#adjust-effort-level) per model for each custom role, alongside role-level [organization model restr
* Levels above the cap aren't offered in the `/effort` picker, and naming a higher level with `--effort` or `/effort` runs at the cap instead. If you start a plain-text `--print` r
* Caps are per model, so switching models can change which levels are available.
+6 −2 lines
Upstream edited 21 Oct 2026 01:33 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
errors
±changed
+125−37
claude-codechanged
errors
| `The browser can't return the sign-in to Claude Code automatically` | [Authentication](#couldnt-start-the-local-sign-in-listener) |
| `Couldn't start the local listener that receives the browser sign-in. Try again.` | [Authentication](#couldnt-start-the-local-sign-in-listener) |
| `The command was not started: this computer's disk is full (ENOSPC)` / `(EDQUOT)` | [Tool errors](#disk-quota-or-temp-filesystem-is-full) |
| `matches no tool · check for typos` / `matches no known tool — check for typos` | [Configuration warnings](#matches-no-tool) |
| `Write(<path>) never matches · use Edit(<path>)` / `Glob(<path>) never matches · use Read(<path>)` | [Configuration warnings](#is-not-matched-by-file-permission-checks) |
+125 −37 lines
Upstream edited 21 Oct 2026 01:33 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
agent-sdk/permissions
±changed
+1−1
claude-codechanged
agent-sdk/permissions
Allow rules accept tool-name globs only after a literal `mcp__<server>__` prefix. The server segment must be glob-free so the rule names a specific server you configured: `mcp__pup
Allow rules accept tool-name globs only after a literal `mcp__<server>__` prefix. The server segment must be glob-free so the rule names a specific server you configured: `mcp__pup
+1 −1 lines
Upstream edited 21 Oct 2026 01:33 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
advisor
±changed
+1−1
claude-codechanged
advisor
If the requested model can act as an advisor but [ranks below](#choose-an-advisor-model) the session's main model, Claude Code starts the session anyway. Outside background session
If the model you pass can act as an advisor but [ranks below your main model](#choose-an-advisor-model), the session starts whether or not it's a background session. Your main mode
+1 −1 lines
Upstream edited 21 Oct 2026 01:33 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
agent-sdk/mcp
±changed
+30−18
claude-codechanged
agent-sdk/mcp
Claude Code registers the servers you pass in `options.mcpServers` at startup and emits the [init message](#error-handling) once the first-turn wait, if any, resolves. Whether each
Claude Code starts connecting your MCP servers when the session starts, and sends the [init message](#error-handling) once the first-turn wait, if any, ends. How long the first tur
For a server in `options.mcpServers`, find its type in the first column to see how long the first turn waits for it.
Servers loaded from [settings files](#from-a-config-file) such as `.mcp.json` or from plugins commonly show `pending` in the init message. When `options.mcpServers` holds a stdio,
+30 −18 lines
Upstream edited 21 Oct 2026 01:30 UTC Recorded here 21 Oct 2026 01:37 UTC
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 21 Oct 2026 01:07 UTC
claude-code
claude-tag
±changed
+4−4
claude-codechanged
claude-tag
No line of the body moved in this read.
+4 −4 lines
Recorded here 21 Oct 2026 00:37 UTC
claude-code
settings-reference
±changed
+2−0
claude-codechanged
settings-reference
Braces match as ordinary characters in a name, so `~/data/{keys,tokens}` covers a directory literally named `{keys,tokens}`, not two alternatives. To cover `~/data/keys` and `~/dat
+2 −0 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
sandboxing
±changed
+4−0
claude-codechanged
sandboxing
#### Denied paths that pass through a symlink
On macOS, when a denied path passes through a symlink, such as a `~/.claude` that is a link into a dotfiles repository, the sandbox also denies the path's real location. This cover
+4 −0 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
claude-apps-gateway-config
±changed
+3−2
claude-codechanged
claude-apps-gateway-config
* **Deny-lists and hook arrays**: `permissions.deny`, `permissions.ask`, `disabledMcpjsonServers`, `deniedMcpServers`, `blockedMarketplaces`, and every `hooks` event-type array. Th
* **Deny-lists and hook arrays**: `permissions.deny`, `permissions.ask`, `disabledMcpjsonServers`, `deniedMcpServers`, `blockedMarketplaces`, `deniedModels`, and every `hooks` even
| [`deniedModels`](/docs/en/settings-reference#deniedmodels), [`availableModelsMatch`](/docs/en/settings-reference#availablemodelsmatch) | Gateway + CLI | [Block specific models or
* The model list, from `availableModels`. [Extended context in Claude Desktop](#extended-context-in-claude-desktop) covers each model's 1M context option
* The model list, from `availableModels`. The list leaves out models that the policy blocks with [`deniedModels` or `availableModelsMatch: exact`](#what-goes-in-cli). This filter r
+3 −2 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
prompt-caching
±changed
+3−1
claude-codechanged
prompt-caching
* **Tools loaded upfront**: adding a definition invalidates the cache, and so does removing one on purpose. This applies when tool search is [below its `auto` threshold, disabled,
* **Tools loaded upfront**: adding a definition invalidates the cache, and so does removing one on purpose. This applies when tool search is [below its `auto` threshold, disabled,
* Google Cloud's Agent Platform models earlier than Opus 4.5, Sonnet 4.5, and Haiku 4.5
* A Microsoft Foundry [deployment hosted on Azure](https://platform.claude.com/docs/en/build-with-claude/claude-in-microsoft-foundry#hosting-options) once Claude Code detects tha
+3 −1 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
mcp
±changed
+19−4
claude-codechanged
mcp
* **Without tool search**: Claude uses the `WaitForMcpServers` tool instead. Configurations without tool search include a custom `ANTHROPIC_BASE_URL`, `ENABLE_TOOL_SEARCH=false`, a
* **Without tool search**: Claude uses the `WaitForMcpServers` tool instead. Configurations without tool search include `ENABLE_TOOL_SEARCH=false` and a model earlier than Claude O
Tool search is enabled by default: MCP tools are deferred and discovered on demand. Claude Code disables it when `ANTHROPIC_BASE_URL` points to a non-first-party host, since most p
Tool search is enabled by default: MCP tools are deferred and discovered on demand.
| (unset) | All MCP tools deferred and loaded on demand. Falls back to loading upfront on models earlier than Opus 4 and Sonnet 4, on Google Cloud's Agent Platform models earlier t
+19 −4 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
llm-gateway-protocol
±changed
+18−7
claude-codechanged
llm-gateway-protocol
Capabilities that add body fields pair them with a beta header, and the pair travels together. A gateway that strips the header while passing the body, or forwards an Anthropic-for
### Beta header and body field pairs
+18 −7 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
llm-gateway-connect
±changed
+1−1
claude-codechanged
llm-gateway-connect
| `400` errors naming `context_management`, `Extra inputs are not permitted`, or other unrecognized fields | The gateway forwards requests to an upstream that rejects fields Claude
| `400` errors naming `context_management`, `Extra inputs are not permitted`, or other unrecognized fields | The gateway forwards requests to an upstream that rejects fields Claude
+1 −1 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
feature-availability
±changed
+1−1
claude-codechanged
feature-availability
* **MCP servers**: [connectors from claude.ai](/docs/en/mcp#use-mcp-servers-from-claude-ai) load only when your claude.ai subscription is the active authentication method. [Tool se
* **MCP servers**: [connectors from claude.ai](/docs/en/mcp#use-mcp-servers-from-claude-ai) load only when your claude.ai subscription is the active authentication method. [Tool se
+1 −1 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
errors
±changed
+15−11
claude-codechanged
errors
A proxy or LLM gateway between Claude Code and the API stripped the `anthropic-beta` request header, so the API rejected fields that depend on it.
```text theme={null}
API Error: 400 ... Extra inputs are not permitted ... context_management
```
+15 −11 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
agent-sdk/tool-search
±changed
+7−3
claude-codechanged
agent-sdk/tool-search
The SDK also disables tool search when `ANTHROPIC_BASE_URL` points to a non-first-party host, since most proxies don't forward `tool_reference` blocks. You can override that defaul
> [!NOTE]
>
> Tool search is on by default behind an `ANTHROPIC_BASE_URL` gateway, with nothing to set. If the gateway rejects it, see [Tool search through an LLM gateway](/docs/en/mcp#tool-se
+7 −3 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
agent-sdk/mcp
±changed
+1−1
claude-codechanged
agent-sdk/mcp
* **With [tool search](/docs/en/agent-sdk/tool-search), the default**: the wait covers still-pending servers configured with [`alwaysLoad: true`](/docs/en/mcp#exempt-a-server-from-
* **With [tool search](/docs/en/agent-sdk/tool-search), the default**: the wait covers still-pending servers configured with [`alwaysLoad: true`](/docs/en/mcp#exempt-a-server-from-
+1 −1 lines
Upstream edited 21 Oct 2026 00:36 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
agent-sdk/examples
±changed
+4−0
claude-codechanged
agent-sdk/examples
## Fork a complete project
To build a shopping agent or a merchant agent with the Python SDK, start from the [Claude for commerce reference implementation](https://platform.claude.com/docs/en/about-claude/us
+4 −0 lines
Upstream edited 21 Oct 2026 00:18 UTC Recorded here 21 Oct 2026 00:37 UTC
claude-code
changelog
±changed
+52−0
claude-codechanged
changelog
**2.1.305**
**Added**
+52 −0 lines
Upstream edited 21 Oct 2026 00:17 UTC Recorded here 21 Oct 2026 00:37 UTC
20 Oct 2026
81 pages
+7391 −8428 lines
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 20 Oct 2026 23:07 UTC
claude-code
self-hosted-environments-deploy
±changed
+9−0
claude-codechanged
self-hosted-environments-deploy
* **Your start script prints a write error containing `Resource temporarily unavailable` and exits**: your script shares its stdout or stderr with the `claude self-hosted-runner` o
To stop the error, send the runner's or orchestrator's output to a file that your script's user can create. Append `>> "$HOME/claude-runner.log" 2>&1` to the command that starts
```bash theme={null}
+9 −0 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
self-hosted-environments-configuration
±changed
+24−0
claude-codechanged
self-hosted-environments-configuration
#### Send a background process's output to a file
If your `spawn-runner` hook leaves a process running in the background, send that process's output to a file that the hook's user can create. For example, append `>> "$HOME/claude-
```bash theme={null}
+24 −0 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
plugins/troubleshooting
±changed
+1−0
claude-codechanged
plugins/troubleshooting
* **`hooks path escapes plugin directory: <path>`**: see [Path escapes plugin directory](/docs/en/errors#path-escapes-plugin-directory)
+1 −0 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
plugins/mods/reference
±changed
+1−1
claude-codechanged
plugins/mods/reference
| `Image` | PNG or RGBA bytes up to 2 MiB, or a file path, `columns` and `rows` up to 255, and `alt` text. | ✓ | | |
| `Image` | PNG or RGBA bytes up to 2 MiB, or a file path, `columns` and `rows` up to 255, and `alt` text. See [which terminals draw it](/docs/en/plugins/mods/gallery#image-and-cli
+1 −1 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
plugins/mods/gallery
±changed
+13−6
claude-codechanged
plugins/mods/gallery
Unless Claude Code detects that the terminal draws kitty graphics protocol images with Unicode placeholders, the user sees an `Image`'s `alt` text, dimmed, in place of the picture.
* **Other terminals**: any terminal that isn't one of those two, or that doesn't answer the query.
* **tmux and screen**: a session running inside tmux or screen, in any terminal, kitty and Ghostty included.
An `Image` draws as a picture in these terminals:
+13 −6 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
errors
±changed
+40−8
claude-codechanged
errors
| `Plugin not loaded: its manifest <path> was not read` | [Plugin errors](#plugin-not-loaded-its-manifest-was-not-read) |
A plugin component path, declared in the plugin's `plugin.json` or in its [marketplace entry](/docs/en/plugins/marketplace-reference#plugin-entries), resolves outside the plugin's
One of the plugin's component paths leads outside the plugin directory or couldn't be confirmed inside it, so Claude Code drops that path and loads the rest of the plugin. For a pa
A message that gives the full path of one of a marketplace plugin's own files, such as `hooks/hooks.json`, means Claude Code skipped that file. That happens when Claude Code can't
+40 −8 lines
Upstream edited 20 Oct 2026 22:43 UTC Recorded here 20 Oct 2026 23:07 UTC
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 20 Oct 2026 22:37 UTC
claude-code
vs-code
±changed
+5−0
claude-codechanged
vs-code
* **Conversations missing with [`claudeProcessWrapper`](#extension-settings) set**: update the extension to version 2.1.304 or later. The list then includes the conversations that
#### See a background subagent's tool calls
When a subagent shows as one status line with no tool calls, type `/tasks` in the prompt box to open the [agent map](#use-the-prompt-box) and click the subagent there. The subagent
+5 −0 lines
Upstream edited 20 Oct 2026 22:35 UTC Recorded here 20 Oct 2026 22:37 UTC
claude-code
artifacts
±changed
+3−2
claude-codechanged
artifacts
Claude Code reads the list from your claude.ai account, so it works in a new session and after `/clear`, when the link has scrolled out of the terminal.
* **`d`**: outside a [cloud session](/docs/en/claude-code-on-the-web), permanently deletes an artifact you own after you confirm. Its link stops working for everyone, including you
Claude Code reads the list from your claude.ai account, so `/artifacts` works in a new session and after `/clear`, when an artifact's link has scrolled out of the terminal.
+3 −2 lines
Upstream edited 20 Oct 2026 22:35 UTC Recorded here 20 Oct 2026 22:37 UTC
claude-code
agent-sdk/typescript
±changed
+27−2
claude-codechanged
agent-sdk/typescript
| `getContextUsage(opts?)` | Returns an [`SDKControlGetContextUsageResponse`](#sdkcontrolgetcontextusageresponse) breaking down the session's context window usage by category, skil
| `getContextUsage(opts?)` | Returns an [`SDKControlGetContextUsageResponse`](#sdkcontrolgetcontextusageresponse) breaking down the session's context window usage by category, skil
Return type of [`getContextUsage()`](#query-object). With the default `detail`, this is the same payload Claude Code renders for the `/context` command in an interactive session, s
Return type of [`getContextUsage()`](#query-object). With the default `detail`, the response includes the payload Claude Code renders for the `/context` command in an interactive s
In `SDKControlGetContextUsageResponse`, `categories[].id`, `mcpInstructions`, and the `pluginId` and `path` fields of `agents` and `skills.skillFrontmatter` entries require Agent S
+27 −2 lines
Upstream edited 20 Oct 2026 22:33 UTC Recorded here 20 Oct 2026 22:37 UTC
claude-code
agent-sdk/python
±changed
+2−2
claude-codechanged
agent-sdk/python
| `get_context_usage()` | Get a breakdown of context window usage by category, skill, and tool. The same data `/context` shows in an interactive session. Returns [`ContextUsageResp
| `get_context_usage()` | Get a breakdown of context window usage by category, skill, and tool. Includes the data `/context` shows in an interactive session. Returns [`ContextUsage
Response from [`ClaudeSDKClient.get_context_usage()`](#methods). This is the same payload Claude Code renders for the `/context` command in an interactive session, so alongside the
Response from [`ClaudeSDKClient.get_context_usage()`](#methods). The response includes the payload Claude Code renders for the `/context` command in an interactive session. Alongsi
+2 −2 lines
Upstream edited 20 Oct 2026 22:33 UTC Recorded here 20 Oct 2026 22:37 UTC
claude-code
claude-tag
±changed
+1−1
claude-codechanged
claude-tag
No line of the body moved in this read.
+1 −1 lines
Recorded here 20 Oct 2026 22:07 UTC
claude-code
artifacts
±changed
+5−5
claude-codechanged
artifacts
| `GET` | `/v1/compliance/code/artifacts` |
| `GET` | `/v1/compliance/code/artifacts/{artifact_id}/versions/{version_id}` |
| `DELETE` | `/v1/compliance/code/artifacts/{artifact_id}` |
For the request and response schemas, see the [Compliance API reference](https://docs.claude.com/en/api/compliance/code/artifacts).
+5 −5 lines
Upstream edited 20 Oct 2026 21:52 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
self-hosted-environments-configuration
±changed
+1−1
claude-codechanged
self-hosted-environments-configuration
On non-zero exit, the tail of the hook's stderr appears in the **Activity** tab as the failure reason, so write the actionable error to stderr and never write secrets there. In
While a session backs off, it shows **Queued** in the **Activity** tab with no spawn error, so read the hook's output in the orchestrator's log. Once blocked, the session shows
+1 −1 lines
Upstream edited 20 Oct 2026 21:48 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
workflows
±changed
+4−2
claude-codechanged
workflows
Claude Code replays the run in the order agents started, and each agent either returns its saved result or runs again:
Claude Code replays the run in the order agents start, and each agent either returns its saved result or runs again:
Before v2.1.305, completed agents could also run again when you relaunched an unedited script whose `pipeline()` items or `parallel()` tasks ran more than one agent each. To stop
* **Failed**: runs again, and so does every agent that started after it, even ones that completed. Stopping one agent alone, by selecting it in [`/workflows`](#watch-the-run) and p
+4 −2 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
vs-code
±changed
+1−1
claude-codechanged
vs-code
* **Status changes**: the extension announces when Claude starts working, when Claude is ready for your input, and when Claude Code starts compacting the conversation.
* **Status changes**: the extension announces when Claude starts working, when Claude is ready for your input, and when Claude Code starts compacting the conversation. If you send
+1 −1 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
sub-agents
±changed
+6−0
claude-codechanged
sub-agents
If a subagent's report opens with the following line, the subagent finished on a fallback model:
```text theme={null}
⚠ This agent switched to <model> because its earlier model could not be used.
```
+6 −0 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
remote-control
±changed
+10−1
claude-codechanged
remote-control
| `--spawn <mode>` | How the server creates sessions. • `same-dir` (default): all sessions share the current working directory, so they can conflict if editing the same files. • `w
| `--spawn <mode>` | How the server creates sessions. • `same-dir` (default when no mode is saved): all sessions share the current working directory, so they can conflict if editin
### Set the default spawn mode for a project
+10 −1 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
plugins/mods/interface
±changed
+2−2
claude-codechanged
plugins/mods/interface
| `columns` | The width to ask for when the pane sits beside the transcript |
| `columns` | The width to ask for when the pane sits beside the transcript, at most 45% of the terminal's width. Once the user has resized a pane beside the transcript, in this se
| Ctrl+X then an arrow key | Resizes your pane. Left or Up gives it more room, and Right or Down gives the room back. |
| Ctrl+X then an arrow key | Resizes your pane. Left or Up gives it more room, and Right or Down gives the room back. A pane beside the transcript stops widening at 45% of the term
+2 −2 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
mcp
±changed
+6−1
claude-codechanged
mcp
For a plugin's stdio server, `claude mcp get` prints `Command: stdio`, an empty `Args:` line, and each environment variable as `NAME=[REDACTED]`. The values are hidden because they
`claude mcp get` hides the values configured for a plugin's stdio or WebSocket server, because they can carry credentials:
* **stdio server**: it prints `Command: stdio`, an empty `Args:` line, and each environment variable as `NAME=[REDACTED]`
* **WebSocket server**: it prints `URL: ws` and each header as `<name>: [REDACTED]`, and `claude mcp list` shows `ws` in place of the URL
+6 −1 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
errors
±changed
+8−4
claude-codechanged
errors
| [`CLAUDE_CODE_RETRY_WATCHDOG_USAGE_LIMIT_RECHECK_MS`](/docs/en/env-vars) | unset | Minimum interval in milliseconds between checks of whether a [usage limit](#youve-hit-your-sess
The default request timeout is 10 minutes.
**What to do:**
+8 −4 lines
Upstream edited 20 Oct 2026 21:42 UTC Recorded here 20 Oct 2026 22:07 UTC
claude-code
vs-code
±changed
+7−1
claude-codechanged
vs-code
The dialog also breaks down what is contributing to your plan limits. It flags behaviors that account for 10% or more of recent usage, such as cache misses, long context, and sub
The dialog also breaks down what is contributing to your plan limits:
* **Recommendations**: changes that lower your usage, one per row with the share of recent usage it addresses, such as `Trim ~/code/api/CLAUDE.md · 49k tokens per request`
* **Fix button**: to have Claude Code apply the recommendations it can, click the button under them, such as **Fix both**. To learn which rows the button acts on and what then ha
+7 −1 lines
Upstream edited 20 Oct 2026 21:32 UTC Recorded here 20 Oct 2026 21:37 UTC
claude-code
skills
±changed
+1−0
claude-codechanged
skills
* **Usage**: on a Pro, Max, Team, or Enterprise plan, the [changes `/usage` recommends](/docs/en/costs#plan-usage-breakdown) to lower your plan usage, measured over the last 7 days
+1 −0 lines
Upstream edited 20 Oct 2026 21:32 UTC Recorded here 20 Oct 2026 21:37 UTC
Nothing on this page matches that. Press Enter to search every page of the ledger.