One read of Claude Code CLIclaude-code-20261009T200701Z
6 pages moved out of 221 read.
Pages moved
6
significant first
Pages read
221
in this capture
Captured
20:07 UTC
Corpus hash
968838c2d1a9
corpus-hash
What this read moved
1-6 of 6changelog Changed · +93 / -0 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 6
66
77Run `claude --version` to check your installed version.
88
9<Update label="2.1.296" description="October 9, 2026">
10 **Added**
11
12 * Added a `code` key to the Claude apps gateway's `managed.policies[]`: the same settings as `cli`, also applied in Claude Desktop's Code tab; beside `desktop`, it turns on Claude Desktop's gateway mode
13 * Added `autoCompactWindow` to subagent frontmatter and `--agents` definitions, so a subagent can auto-compact earlier than the main conversation's window
14 * Added `CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL` to run every workflow agent on one model while other subagents keep theirs
15 * Added `CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS` environment variable to set a longer maximum delay for the backoff when retrying an overloaded (529) request
16 * Added a note in `/plugin` on a plugin whose hooks are left out because another enabled plugin has the same name
17 * Added an `allow_large` option to the Read tool so Claude can read a text file past the usual size limits in one call when it needs the whole file and the context has room
18 * \[Cloud sessions] Added a status filter to the Sessions and Runners lists on a self-hosted environment's Activity tab in admin settings; you can pick several statuses at once
19 * \[Claude Tag] Added creating, editing and deleting workspace and channel memory files on the Activity page's Memory tab for members with the Claude Tag Admin permission
20
21 **Changed**
22
23 * Updated `/cost`, the status line, `--max-budget-usd` and the SDK's cost figures to price Sonnet 5.5 cache reads at $0.10 per million tokens (was $0.20)
24 * Updated the bundled dataviz skill: lighter violet for the seventh series in light mode, softer primary text in dark mode, and compact y-axis labels like 1K
25 * Changed the default limit on MCP tool descriptions sent up front and on MCP server instructions from 2,048 to 4,096 characters
26 * Changed `←`: a turn or `!` command started while the session moves to the background is now stopped instead of finishing out of sight
27 * \[VSCode] Changed Claude in Chrome to ask before browser actions in every session, including one connected with `@browser`, as the terminal does; allowing a site for the session stops repeat asks
28 * \[Claude Tag] Changed Claude in Slack to show the footer with the session link, model and cost only on its newest reply in a thread; the reply before it drops the footer
29
30 <Expandable title="9 improvements">
31 - Improved fullscreen mode: the link under the mouse pointer is now underlined to show it can be opened
32 - Improved transcript toggle (ctrl+o) responsiveness in code-heavy conversations by caching syntax-highlighted code blocks
33 - Improved Claude Desktop's Code tab behind a Claude apps gateway: when a session can't start, including on a machine not set up for the gateway's `code` settings, the reason shows as a reply
34 - Improved the errors for a refused cloud session: they now say why your organization's policy couldn't be loaded, and name an API key or auth token that overrides your claude.ai login
35 - Improved `--debug` output to name unrecognized frontmatter fields in custom agent files, with a hint for likely typos
36 - Improved auto mode: a tool call that wasn't run because auto mode's check had no usable answer now shows as a dim "Not run" row instead of a red error
37 - Improved `--debug` output for hooks: command hooks on tool calls, prompts, SessionStart and Stop now log their command, plugin, outcome and duration when they finish, so a slow hook is identifiable
38 - Improved `CLAUDE_CODE_TRANSCRIPT_LOCAL_GC`: a large transcript file is no longer rewritten when that would free under 10% of it
39 - \[Code Review] Improved Code Review admin settings to say why a setting couldn't be saved, such as a policy restriction, instead of a generic failure message
40 </Expandable>
41
42 <Expandable title="56 bug fixes">
43 * Fixed managed-settings `PreToolUse` hooks that deny a tool call with `"continue": false`, and managed `prompt` hooks that block one, refusing the call but not ending the turn
44 * Fixed PostToolUse hooks in managed settings not applying `updatedMCPToolOutput` in some sessions
45 * Fixed headless sessions starting a folder's `.mcp.json` or plugin MCP server that was switched off for that folder, after changing directory or reloading plugins
46 * Fixed a Claude apps gateway that serves `allowedProviders` with `"gateway"` locking out laptops that name that gateway in user settings
47 * Fixed a saved Claude apps gateway sign-in being ignored on machines whose managed settings set `forceLoginMethod` to `gateway` with no `forceLoginGatewayUrl` (regression in 2.1.295)
48 * Fixed `--teleport` opening an empty conversation when the session's history could not be read
49 * Fixed token counts for Haiku 5.5 and other models that take only adaptive thinking, which failed behind some gateways and were counted with budget thinking elsewhere
50 * Fixed resumed subagents being told that the user rejected a tool call that a session shutdown had interrupted
51 * Fixed hook output being altered when it contained text resembling a plugin hint tag
52 * Fixed secret redaction in shared transcripts and debug logs missing some values that follow a key with no value, including in JSON written inside a shell string
53 * Fixed a stray `52;c;…` escape sequence printed on screen after copying in older VTE-based terminals such as MATE Terminal
54 * Fixed toasts and notifications waiting unseen for as long as the `/diff` panel or dialog was open
55 * Fixed Esc or an interrupt during a `UserPromptSubmit` hook or a mod's `prompt.submit` hook ending headless sessions, clearing the typed prompt, or letting the unchecked prompt through
56 * Fixed SessionStart hooks of a plugin loaded after a headless session starts being skipped when a different plugin with the same name, or another spelling of it, had already run
57 * Fixed `claude self-hosted-runner` printing misleading errors when registration is refused: it now names the org admin setting, or says a restarted on-demand runner needs a fresh work order
58 * Fixed a self-hosted runner session's git fetch, pull and push failing while another session of the same repository starts on a runner with `--capacity` above 1
59 * Fixed workflow scripts silently truncating results, agent options and other values nested deeper than about 20,000 levels, and deeply nested results writing huge task output files
60 * Fixed a plugin's served `$` methods running in the main session's working directory instead of the calling agent's, and ignoring the turn their calling hook holds
61 * Fixed `$.agent.register` succeeding from a mod's hook that was still running after the mod was reloaded or removed; the call is now refused
62 * Fixed `$.http.fetch` in mods refusing a `HEAD` request when the response declares a `Content-Length` over the 4 MiB body limit
63 * Fixed `claude plugin marketplace add`, `marketplace update` and `plugin install` failing with an internal error for a marketplace named like `constructor`; `add` now refuses such names clearly
64 * Fixed a plugin secret named `constructor` or `prototype` being deleted by the next save of that plugin's options
65 * Fixed cloud sessions skipping auto mode's checks on Claude in Chrome actions that a saved permission allows when their feature flags had not loaded
66 * Fixed `CLAUDE_CODE_RESUME_INTERRUPTED_TURN` re-running a finished turn after a restart when an MCP tool result had ended that turn
67 * Fixed Bash permission checks auto-approving some commands that assign the `BASH_ARGV0` shell variable and then use it; these now prompt for approval
68 * Fixed Edit and NotebookEdit replacing every non-ASCII character in files that are not valid UTF-8 (Windows-1252, Shift-JIS, GBK); such edits are now refused
69 * Fixed a prompt sent just after `←` being run twice, once unseen in the foreground, when the background service was slow to answer
70 * Fixed a prompt sent while SessionStart hooks were still running vanishing when `←` moved the session to the background; it is still not sent, but `↑` now brings it back
71 * Fixed a late-loading plugin's SessionStart hook output reaching the new conversation after `/clear` in headless sessions
72 * Fixed PowerShell commands not seeing variables that hooks write to CLAUDE\_ENV\_FILE (when the file holds only plain assignments)
73 * Fixed Claude calling cloud sessions non-interactive and suggesting `/mcp` or `claude mcp` when an MCP server needs authentication
74 * Fixed `/code-review` ending on a raw JSON array in cloud sessions, the Agent SDK and IDE integrations; the findings now print as a numbered list
75 * Fixed auto mode occasionally refusing one update to your own artifact, saying its sharing had changed when it had not
76 * Fixed skills synced from claude.ai never finishing their install, and being downloaded again at every sync, when several sessions share a config directory
77 * Fixed a plugin synced from claude.ai being disabled when the marketplace dependency it declares was also synced from claude.ai
78 * Fixed a startup hang on macOS when plugin paths copied from a Linux machine pointed under /home
79 * Fixed the Workflow tool refusing script files with CRLF line endings, such as files checked out on Windows
80 * Fixed OpenTelemetry `tool_result` events missing `git_commit_id` after `git commit -q` or `git -C <dir> commit`
81 * Fixed `CLAUDE_CODE_TRANSCRIPT_LOCAL_GC` dropping a message from the saved transcript when its text held a Unicode line or paragraph separator (U+2028, U+2029)
82 * Fixed `claude purge` leaving a prompt in the history file when its text held a Unicode line or paragraph separator (U+2028, U+2029)
83 * Windows: Fixed PowerShell commands longer than about 1 KB always asking for permission, so allow rules and read-only detection now apply up to 32 KB
84 * Windows: Fixed stdio MCP servers being force-killed at shutdown; stdin now closes first and the process tree is killed only if the server is still running after 300 ms
85 * Windows: Fixed `claude plugin install` failing for GitHub `owner/repo` plugin sources on machines with no GitHub SSH key; the clone is now retried over HTTPS
86 * Windows: Fixed `rm -rf /c/Users/<name>` in Git Bash not asking in bypass permissions mode
87 * \[VSCode] Fixed `claudeCode.spinnerVerbs` missing from the extension's settings, so settings.json now completes and validates it; a malformed value no longer breaks the chat panel
88 * \[VSCode] Fixed chat links written as a full Windows path, such as C:\repo\file.ts or file:///C:/repo/file.ts, not opening the file on Windows
89 * \[VSCode] Fixed the working indicator still animating when reduced motion is turned on in your OS, in VS Code, or with the `prefersReducedMotion` setting
90 * \[VSCode] Fixed the prompt cache clock reading warm after the panel caught up on messages late, such as after the display was off or a remote window reconnected
91 * \[Cloud sessions] Fixed every cloud session waiting for a self-hosted runner showing the same queue position; each now shows how many sessions are ahead of it
92 * \[Cloud sessions] Fixed a message you send while Claude is still replying in a cloud session jumping above that reply when it finishes; it now stays below it, including after a reload
93 * \[Claude Tag] Fixed links to URLs that contain parentheses breaking in some of Claude's Slack replies and task checklists, which showed stray bracket text and opened the wrong page
94 * \[Claude Tag] Fixed Claude's Slack replies drawing color swatches beside codes that only look like hex colors, such as six-digit pull request numbers
95 * \[Claude Tag] Fixed the plugin confirmation card Claude posts in a Slack channel showing a plugin's ID instead of its name
96 * \[Claude Tag] Fixed Claude in Slack posting a card to remove a plugin the channel's own settings don't include; it now explains where the plugin comes from instead
97 * \[Claude Tag] Fixed Claude in Slack sometimes losing its reply when a session in a self-hosted environment moved to a new runner partway through a task
98 * \[Code Review] Fixed the Code Review check sometimes staying "in progress" on a pull request after the review finished; it now shows the review's result
99 </Expandable>
100</Update>
101
9102<Update label="2.1.295" description="October 8, 2026">
10103 **Added**
11104
from line 1894
18011894 * Fixed a `sandbox.excludedCommands` glob exempting an entire compound Bash command from the sandbox when only one part matched; every part must now match
18021895 * Fixed resumed subagents and teammates re-rendering the MCP tool definitions they had loaded, which broke prompt caching for that agent
18031896 * Fixed rate-limited artifact publishes telling Claude to stop retrying; Claude is now told nothing was published and when to send the same publish again
1804 * Fixed attachments recorded earlier in a conversation being re-rendered after a resume or relaunch, which dropped extended thinking and missed the prompt cache
1805 * Fixed Console sign-in showing only "Request failed with status code 400" when the server refuses to create an API key; it now shows the server's message
1806 * Fixed messages typed while Claude is still working sometimes being ignored by the model
1807 * \[VSCode] Fixed the "General config" menu row showing `/config` usage text instead of opening settings, and made typed `/mcp`, `/hooks`, `/memory`, `/rewind` and similar commands open their dialogs
1808 * \[VSCode] Fixed the effort slider's level not persisting into later sessions on a model that already had a level saved with `/effort`
1809 * \[VSCode] Fixed Auto missing from the mode picker for conversations opened in an already-used panel when the saved model setting is a differently-cased alias such as "Sonnet"
1810 * \[VSCode] Fixed `/fast` not saving fast mode as the default, so it was lost when the extension relaunched Claude Code
1811 * \[Claude Code on the web] Fixed a cloud environment saved with Custom network access and no domains silently reverting to Trusted; the dialog now asks for at least one domain
1812 * \[Claude Tag] Fixed routines created in a Slack channel on an Enterprise Grid org-wide install failing to read other public channels in their workspace when they ran
1813 * \[Claude Tag] Fixed the "Learn more" links on credential presets in Claude Tag access bundles to open each vendor's credential-setup page instead of a generic API reference
1814 * \[Claude Tag] Fixed Google Cloud credential forms in Claude Tag access bundles: a refused key file now says why, the website and scopes stay locked, and a rejected rotation keeps the pasted key
1815 * \[Claude Tag] Fixed the network events log in Claude Tag admin settings showing no response status for requests through connections that use AWS signing, client certificates or a custom CA
1816 </Expandable>
1817</Update>
1818
1819<Update label="2.1.276" description="September 18, 2026">
1820 * Fixed every request failing with `400 … Input tag 'advisor_20260301'` when `ANTHROPIC_BASE_URL` points at a proxy or gateway (2.1.275 regression)
1821</Update>
1822
1823<Update label="2.1.275" description="September 17, 2026">
1824 **Added**
1825
1826 * Added the signed-in account to Claude apps gateway sign-in: when the gateway names it, you confirm it before the credential is saved, and `/status` shows it
1827 * Added a send-now key (ctrl+enter, or ctrl+x ctrl+s) that interrupts the current turn and sends all queued messages at once; sent and queued messages show in gray until the model receives them
1828 * Added a startup warning when a configured `otelHeadersHelper` fails, so sessions that silently export no telemetry are noticed
1829 * Added syncing of the skills and plugins enabled on your claude.ai account to terminal sessions signed in with it; opt out with `syncClaudeAiSkills: false` or `syncClaudeAiPlugins: false`
1830 * Added `/plugin install <plugin> --marketplace <source>`, which offers to add the marketplace before installing the plugin
1831 * \[VSCode] Added viewing, editing and deleting a saved memory inside the Memory dialog
1832 * \[VSCode] Added sending an attached image without typing any text
1833 * \[VSCode] Added a Retry link to the MCP servers dialog when the server list fails to load
1834 * \[VSCode] Added accept and reject buttons under each change in the proposed-change diff tab, so an edit can be reviewed change by change
1835 * \[Claude Code on the web] Added a "New routine" button to the page shown when a routine link no longer resolves, next to the link back to your routines list
1836 * \[Claude Tag] Added attach conditions for access bundles in Claude Tag settings: an Owner can let a bundle also apply in channels with guests or Slack Connect channels, not just member-only
1837 * \[Claude Tag] Added Amazon CloudWatch, CloudWatch Logs, Amazon SNS, Google Cloud Monitoring and Cloud Logging presets to an access bundle's Credentials tab in Claude Tag admin settings
1838 * \[Claude Tag] Added Datadog presets for the US3, AP1, AP2 and US1-FED sites; new Datadog connections are now limited to Datadog's read and query API routes
1839
1840 **Changed**
1841
1842 * Updated `/logout` for Claude apps gateway sign-ins to also end the session on gateways that advertise token revocation
1843 * Changed hosted sessions to keep an unanswered permission prompt up after a container restart, instead of asking again
1844 * Changed the Artifact tool to ask for a one-word tab icon on a first publish instead of an emoji favicon
1845 * Changed Claude in Chrome in auto mode to skip the extension's per-site check for classifier-approved calls, as bypass mode does, fixing `browser_batch` "Permission denied" after a redirect
1846 * Changed plugins installed from an npm source to be fetched with `npm pack --ignore-scripts` and integrity-verified, so a package's install scripts no longer run
1847 * Changed scheduled and Run now routine runs to save data to, and republish the page of, an artifact you can edit without asking; public artifacts, first publishes and deletes still ask
1848 * Removed the startup notice that told you a one-off scheduled routine had run since your last session
1849 * \[VSCode] Changed New session in a Claude editor tab to open in the sidebar when Preferred Location is set to Sidebar, instead of always opening another tab
1850 * \[VSCode] Changed a message sent while Claude is working to wait at the bottom of the conversation until Claude starts on it
1851
1852 <Expandable title="13 improvements">
1853 - Improved prompt caching for a `--system-prompt` that contains a `__SYSTEM_PROMPT_DYNAMIC_BOUNDARY__` line: the text above it is now cached globally, as the SDK's array form already is
1854 - Improved the `/desktop` error when Claude Desktop does not open: it now says why and what to do next
1855 - Improved the Artifact tool's publish and read results: they now say who can open the page and what the owner's Share menu offers
1856 - Improved artifact publish results: they name the tab icon sent, warn when the page contains a NUL byte, and retry a flaky fetch of the newer page to merge after a stale publish
1857 - Improved pasted and attached images: they are now saved where Claude can open them as files without a permission prompt, including in Desktop and VS Code
1858 - Improved the Artifact tool's guidance so Claude updates a shared artifact in place when you were given edit access to it, instead of publishing a separate copy
1859 - Improved plan-usage reads: editor windows and non-interactive sessions on one machine now share a read made in the last minute instead of each calling the usage endpoint
1860 - Improved the `ListPlugins` tool description so Claude knows it lists plugins enabled on your claude.ai account, not plugins installed locally with `/plugin`
1861 - Improved responsiveness when the terminal is slow or paused: output no longer falls further behind while the terminal catches up
1862 - Improved Write and Edit results for files in the synced account-skills folder: they now say the change is not saved to your account and how to save it
1863 - \[VSCode] Improved the agent map: the pill counts running agents and turns red after a failure, the main agent stays in view while the map scrolls, and agents sort by state then end time
1864 - \[Claude Code on the web] Improved what Claude tells you when asked to edit, delete or run a routine it didn't create: it now links to the routine's page so you can do it yourself
1865 - \[Claude Tag] Improved Claude's notices in Enterprise Grid channels shared across workspaces: they now say when no workspace is set up yet, or why only organization defaults apply
1866 </Expandable>
1867
1868 <Expandable title="60 bug fixes">
1869 * Fixed a restored memory file's age note changing between requests after a compaction or resume, which caused prompt cache misses
1870 * Fixed `--forward-subagent-text` stream-json and SDK output dropping the messages of subagents spawned by a `context: fork` skill, and of forked skills invoked by a subagent or another forked skill
1871 * Fixed @-mention file suggestions being buried below MCP resources when using a custom `fileSuggestion` command or typing `@.`/`@./`
1872 * Fixed fullscreen mode placing background-task completion notices beneath a long turn's collapsed tool row instead of where they arrived; each notice now closes the open row
1873 * Fixed `claude plugin marketplace update` deleting a GitHub marketplace's local copy when the fetch failed and the marketplace was named after its repository
1874 * Fixed plugin and marketplace messages, logs and `claude plugin marketplace list` showing a password or token stored in a git, ssh or marketplace URL
1875 * Fixed a resumed cloud session leaving an unanswered question open in the transcript after a queued message superseded it
1876 * Fixed vim mode placing the cursor one character right after a dot-repeated "!" or a fast-typed "i!" switched a non-empty prompt into shell mode
1877 * Fixed fullscreen mode freezing or blanking for several seconds when scrolling up past a large file diff
1878 * Fixed a stray `</ccmemory>`-style closing tag occasionally appearing in responses
1879 * Fixed plugin messages, logs and the VS Code plugin dialog showing the wrong server for some git addresses
1880 * Fixed a terminal `API Error: 400` on every turn for users behind a network gateway that rewrites API error responses when a beta request header is rejected
1881 * Fixed sandboxed Bash commands on Linux reporting exit code 0 for failed commands when the shell is zsh
1882 * Fixed the Read tool hanging instead of reporting an error when part of a large file could not be decoded under memory pressure
1883 * Fixed `--resume`, the resume picker preview, resumed background agents and the transcript view failing on a session whose saved history contains a malformed task-reminder or @-file attachment entry
1884 * Fixed a crash when resuming a conversation whose transcript contains a malformed message entry, and a fullscreen crash when such a conversation received new messages while scrolled up
1885 * Fixed sessions failing to resume or start when their saved transcript contains a malformed message content block
1886 * Fixed Grep, Glob and @-file suggestions hanging or running out of memory on searches over the 20MB output cap, and system ripgrep reporting "no matches" instead of an error after a flood of warnings
1887 * Fixed `/rewind` in a forked or background session restoring a zero-filled or truncated file when the session's file-history backups could not be fully copied
1888 * Fixed fullscreen sessions sometimes exiting with "Claude Code exited after an unrecoverable interface error" when typing fast or holding a key with the slash-command dropdown open
1889 * Fixed background sessions crashing and restarting their worker when a command fed through stdin ran on a machine that had run out of file descriptors
1890 * Fixed a crash at launch when `~/.claude.json` holds a malformed `mcpNeedsAuthNoticed` value
1891 * Fixed `--resume` and `--continue` dropping a conversation's earlier thinking when a built-in tool it started with has since been switched off by a server-side flag
1892 * Fixed text selected with the mouse in the fullscreen `claude --resume` session picker never reaching the clipboard
1893 * Fixed plugin reload previews replacing a running session's extracted plugin files when the plugin was loaded from a `--plugin-dir` or `--plugin-url` archive
1894 * Fixed self-hosted runners with `--drain-wait-sec` losing the final result of a turn that finished during a SIGTERM drain; the runner now waits briefly for the turn to be reported
1895 * Fixed `SubagentStop` hooks with a specific `matcher` firing for every stopping subagent whose agent type was empty
1896 * Fixed sandboxed Bash commands being unable to write to project directories named `hooks/` or `config/`
1897 * Fixed Artifact updates failing with "File not found" after a session resumes on another machine or its scratchpad is cleared: the page's last published version is restored
1898 * Fixed `/update-config` writing `Write(path)` permission rules, which file permission checks don't match, instead of `Edit(path)` rules
1899 * Fixed four dead documentation URLs (Pricing, Computer Use, Skills, CLI) in the bundled claude-api skill's live-sources table
1900 * \[VSCode]
1897 * Fixed attachments recorded earlier in a conversation being re-rendered a
plugins/mods/api Changed · +89 / -8 lines
### Send one prompt ### Use prompt caching #### Choose between `prompt` and `system` #### Check for cache hits ### What a `model.complete` hook receives
from line 67
6767
6868## Call a model
6969
70A mod can ask a model a question of its own, outside the conversation, for a small job such as sorting or summarizing a piece of text. `$.model.complete` sends one prompt to a model with your session's credentials and resolves to the reply. It has no conversation history.
70A mod can send its own requests to a model for a small job such as classifying or summarizing text. `$.model.complete` sends your prompt on its own, and `$.model.fork({ prompt })` sends the current conversation with your prompt at the end.
7171
72This table compares what each request contains:
73
74| In the request | `$.model.complete` | `$.model.fork` |
75| :- | :- | :- |
76| Model | The `model` you pass | The session's model |
77| System prompt | A short [attribution block](/docs/en/llm-gateway-protocol#system-prompt-attribution-block), then your `system` if you pass one | The session's system prompt |
78| Messages | One user message, your `prompt` | The conversation so far, then your `prompt` as a user message |
79| CLAUDE.md and other project context | Not included | Included, as in the conversation's last request |
80| Tools | None | Claude's tools, which the model can't call |
81
82A fork repeats the conversation's last request, so the Claude API serves most of it from the [prompt cache](/docs/en/prompt-caching) while the conversation is still cached.
83
84Both calls use the session's credentials, so they bill to the user's plan, API key, or cloud provider. [The types for your build](/docs/en/plugins/mods/create#get-the-types-for-your-build) document every `$.model` method.
85
86### Send one prompt
87
88Pass `model` and `prompt` to `$.model.complete`. `prompt` becomes the user message. To give the model instructions, such as a role or an output format, also pass `system`, which becomes the system prompt.
89
7290This hook answers a `/triage` command, [registered as a command](#add-a-command), by asking a small model to label the text typed after it:
7391
7492```javascript theme={null}
from line 106
88106})
89107```
90108
91When you run `/triage the export button does nothing`, the mod sends that text to the model and prints its answer, such as `Label: bug`. Claude's conversation isn't part of the request. When the model doesn't answer, the label is `unknown`.
109When you run `/triage the export button does nothing`, the mod sends that text to the model and prints its answer, such as `Label: bug`. When the model doesn't answer, the label is `unknown`.
92110
93A Claude API failure doesn't reject the call, so check `r.isAnswered`, and read `r.reason` when it's `false`. The call rejects for a request Claude Code won't send, such as a model your organization blocks. [The types for your build](/docs/en/plugins/mods/create#get-the-types-for-your-build) list the other options, such as `effort`, and the [limits](/docs/en/plugins/mods/reference#limits) give the `maxTokens` default.
111A Claude API failure doesn't reject the call, so check `r.isAnswered`, and read `r.reason` when it's `false`. The call rejects for a request Claude Code won't send, such as a model your organization blocks.
94112
95`$.model.fork({ prompt })` asks one question over the current conversation instead, with the same model and system prompt, so the Claude API serves most of it from the prompt cache.
113[The types for your build](/docs/en/plugins/mods/create#get-the-types-for-your-build) list the other options, such as `effort`, and the [limits](/docs/en/plugins/mods/reference#limits) give the `maxTokens` default.
96114
97These calls use the user's plan or API key.
115### Use prompt caching
116
117`$.model.complete` supports the Claude API's [prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching). The API caches the start of a request, called the prefix, up to a [cache breakpoint](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#explicit-cache-breakpoints) that you set. When every call starts with the same long static content, such as instructions or reference material, set a breakpoint at the end of that content. Later calls then read it from the cache instead of paying the full input price for it.
118
119To set a breakpoint, pass `prompt` as an array of `{ text }` blocks instead of a string, and add `cache: true` to the last block of the static content. Claude Code sends that block with the API's `cache_control` field. `system` takes the same array form. To decide between them, see [Choose between `prompt` and `system`](#choose-between-prompt-and-system).
120
121<Note>
122 Arrays of blocks require Claude Code v2.1.292 or later. Earlier versions reject an array in `prompt` with an error that ends with `takes { model, prompt } (host check)`, and they leave an array in `system` out of the request.
123</Note>
124
125This version of the [`/triage` hook](#send-one-prompt) sends a long set of labeling rules before the text to label, with a breakpoint after the rules. `RULES` is a string of your own:
126
127```javascript theme={null}
128on('command.run', { command: 'triage' }, async ($, e) => {
129 const r = await $.model.complete({
130 model: 'haiku',
131 prompt: [
132 // Identical on every call, so it forms the cached prefix
133 { text: RULES, cache: true },
134 // Changes on every call, so it goes after the breakpoint
135 { text: e.args },
136 ],
137 })
138 return { text: 'Label: ' + (r.isAnswered ? r.text.trim() : 'unknown') }
139})
140```
141
142The TTL and the number of breakpoints have these limits:
143
144* **TTL**: a cache entry lasts five minutes after its last use. The TTL comes from the user's Claude Code settings, not from the call. For one hour, set [`subagentPromptCacheTtl`](/docs/en/prompt-caching#choose-the-ttl-yourself) to `1h`.
145* **Breakpoints per request**: the API accepts [up to four](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#when-to-use-multiple-breakpoints), and one more comes back as an `api-error` in `r.reason`
146
147#### Choose between `prompt` and `system`
148
149Put the static content that your calls share at the start of `prompt` unless you know that your requests go directly to the Claude API:
150
151* **Directly to the Claude API, with an API key or a Claude subscription**: either field works
152* **Through [Amazon Bedrock](/docs/en/amazon-bedrock), [Claude Platform on AWS](/docs/en/claude-platform-on-aws), [Google Cloud's Agent Platform](/docs/en/google-vertex-ai), [Microsoft Foundry](/docs/en/microsoft-foundry), or an [LLM gateway](/docs/en/llm-gateway)**: use `prompt`. Claude Code begins the system prompt with an [attribution block](/docs/en/llm-gateway-protocol#system-prompt-attribution-block) whose fingerprint comes from the start of the user message. The `api.anthropic.com` endpoint strips that block before caching. Other endpoints receive it as part of the prompt, so a breakpoint in `system` can miss when `prompt` starts differently.
153* **In a mod that other people run**: use `prompt`, because you don't choose their provider
154
155`system` comes before `prompt` in the prefix, so a breakpoint in `prompt` covers `system` too, and a call with a different `system` misses the cache.
156
157#### Check for cache hits
158
159The result of `$.model.complete` has a `usage` object with the API's [cache fields](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#tracking-cache-performance). `usage.cache_creation_input_tokens` counts the tokens the call wrote to the cache, and `usage.cache_read_input_tokens` counts the tokens it read from the cache. Expect a write on the first call and reads on later calls within the TTL.
160
161If every call writes and none reads, the prefix differs between calls or the calls are further apart than the TTL. For a prefix that differs, see [Choose between `prompt` and `system`](#choose-between-prompt-and-system).
162
163If both fields stay at zero on calls the model answered, nothing was cached. Check for each of these causes:
164
165* **The prefix is too short**: the API doesn't cache a prefix under the model's [minimum length](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#cache-limitations), and it returns no error
166* **Prompt caching is disabled**: when a [`DISABLE_PROMPT_CACHING` variable](/docs/en/prompt-caching#disable-prompt-caching) applies to the model, Claude Code removes the breakpoints and sends the text uncached
167* **Your gateway strips `cache_control`**: a gateway can [remove the field and still return success](/docs/en/prompt-caching#where-the-cache-lives)
168* **Another mod rewrites the start of the text**: Claude Code then [sends it without breakpoints](#what-a-model-complete-hook-receives)
169
170### What a `model.complete` hook receives
171
172If you hook the [`model.complete`](/docs/en/plugins/mods/reference#mods-api-calls) event to inspect or change other mods' requests, read the text from these fields:
173
174* **`e.prompt`**: always a string. When the caller passed an array, it's the blocks' text concatenated in order.
175* **`e.system`**: a string built the same way, or absent when the caller passed no `system`
176* **`e.promptBlocks` and `e.systemBlocks`**: the caller's arrays, each present when the caller passed an array for that field
177
178Claude Code sends the strings your hook passes to `next` and uses the arrays you pass with them to place [cache breakpoints](#use-prompt-caching). It keeps the leading blocks that still match the start of the string, with their breakpoints, and sends the rest of the string with no breakpoint. For example, `next({ ...e, prompt: e.prompt + NOTE })` keeps the caller's breakpoints, and a hook that changes the start of `prompt` removes them.
98179
99180## Run work in the background
100181
claude-code-on-the-web Changed · +1 / -1 lines
from line 417
417417Reopen the session from [claude.ai/code](https://claude.ai/code) to provision a fresh VM:
418418
419419* **Restored**: your conversation history
420* **Not restored**: background work that was still running when the VM was reclaimed, such as subagents and shell commands
420* **Not restored**: background work that was still running when the VM was reclaimed, such as subagents and shell commands, and the pending wakeup of a [self-paced `/loop`](/docs/en/scheduled-tasks#let-claude-choose-the-interval). To restart the loop, run `/loop` again.
421421
422422## Limitations
423423
claude-tag Changed · +1 / -1 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
Nothing in the body moved in this read. What changed is above.
monitoring-usage Changed · +2 / -0 lines
from line 1090
10901090
10911091Logged when Claude Code resolves an `@`-mention in a prompt. Not every mention emits an event: early-exit paths such as permission denials, oversized files, PDF reference attachments, and directory listing failures return without logging.
10921092
1093Each time Claude Code reads a prompt, it logs at most 100 events with a `mention_type` of `"agent"` and 100 with `"mcp_resource"`. Mentions past either limit still resolve but emit no event.
1094
10931095**Event Name**: `claude_code.at_mention`
10941096
10951097**Attributes**:
vs-code Changed · +1 / -0 lines
from line 154
154154* **Bookmarks**: hover over a response and click **Bookmark response** to save it, or click **Remove bookmark** on a saved response to remove it.
155155
156156 To review saved responses, open the Bookmarks panel: click the bookmark icon at the top of the Claude Code panel, select **Bookmarks** in the Context section of the command menu, or type `/bookmarks`. Requires Claude Code v2.1.286 or later.
157* **Files Claude sends you**: when the session is connected to [Remote Control](/docs/en/remote-control#start-a-remote-control-session) and Claude sends you files with the [`SendUserFile` tool](/docs/en/tools-reference), the conversation shows a row such as **Sent report.md, chart.png**. Click a file's name to open it in the editor.
157158* **Context indicator**: the prompt box shows how much of Claude's context window you're using. Claude automatically compacts when needed, or you can run `/compact` manually.
158159* **Prompt cache clock**: a clock icon next to the context indicator estimates how much time the conversation's [prompt cache](/docs/en/prompt-caching) has left before it expires. It counts down from the cache's five-minute or one-hour [lifetime](/docs/en/prompt-caching#cache-lifetime), and each response that uses the cache restarts the countdown. Apart from compaction, the [actions that invalidate the cache](/docs/en/prompt-caching#actions-that-invalidate-the-cache) don't reset the clock, so it can still show minutes left after you switch models.
159160 * Until the countdown runs out, the icon shows the minutes left, such as **12m**.