One read of Claude Code CLIclaude-code-20260926T003701Z
20 pages moved out of 210 read.
Pages moved
20
significant first
Pages read
210
in this capture
Captured
00:37 UTC
Corpus hash
bc6171f49b40
corpus-hash
What this read moved
1-20 of 20cloud-environments Changed · +14 / -3 lines
### Time limits
from line 241
241241
242242## What's available in cloud sessions
243243
244In Anthropic-hosted environments, each session gets a fresh virtual machine (VM) running Ubuntu 24.04 on x86\_64, regardless of your own operating system and CPU architecture, with your repository cloned and common toolchains pre-installed. When a dependency provides precompiled binaries, such as Ruby gems with native extensions or prebuilt Python wheels, use its x86\_64 Linux build to match the VM. This section covers the Anthropic-hosted defaults, the built-in GitHub tools, how to [run tests and services](#run-tests-start-services-and-add-packages), and the [resource limits](#resource-limits) each VM gets.
244In Anthropic-hosted environments, each session gets a fresh virtual machine (VM) running Ubuntu 24.04 on x86\_64, regardless of your own operating system and CPU architecture, with your repository cloned and common toolchains pre-installed. When a dependency provides precompiled binaries, such as Ruby gems with native extensions or prebuilt Python wheels, use its x86\_64 Linux build to match the VM. This section covers the Anthropic-hosted defaults, the built-in GitHub tools, how to [run tests and services](#run-tests-start-services-and-add-packages), the [resource limits](#resource-limits) each VM gets, and the [time limits](#time-limits) on long-running work.
245245
246246<Note>
247247 Sessions your organization routes to a [self-hosted environment](/docs/en/self-hosted-environments) run on your own runners instead, with the tools your runner image provides.
from line 363
363363
364364The VM may stop tasks that need significantly more memory, such as large build jobs or memory-intensive tests. For workloads beyond these limits, use [Remote Control](/docs/en/remote-control) to run Claude Code on your own hardware, or run cloud sessions in a [self-hosted environment](/docs/en/self-hosted-environments) on compute your organization operates.
365365
366### Time limits
367
368In Anthropic-hosted environments, these time limits apply to long-running work in a cloud session, such as a build, an install, or a test run. Each entry links to the section that defines the limit.
369
370* **Commands Claude runs**: a cloud environment doesn't set its own command timeout, so the Bash tool's defaults apply. Claude waits 2 minutes for a command by default and can ask for up to 10 minutes. When a command reaches its [timeout](/docs/en/tools-reference#timeout-and-output-limits), Claude Code [moves it to the background](/docs/en/tools-reference#background-commands) instead of stopping it, unless the command starts with `sleep`.
371* **SessionStart hooks**: Claude Code cancels a `command` hook after 600 seconds unless you set [`timeout`](/docs/en/hooks#common-fields), in seconds, on the hook entry. Claude Code doesn't enforce the timeout on a hook you run with [`async: true`](/docs/en/hooks#run-hooks-in-the-background).
372* **Setup script**: a script that takes longer than roughly five minutes isn't cached. [Script requirements](#script-requirements) covers how to stay under that.
373* **Idle sessions**: a session stops after a period of inactivity and its VM is reclaimed. [Environment expired](/docs/en/claude-code-on-the-web#environment-expired) covers what counts as inactive and how to reopen the session.
374
375To raise the command timeouts for an environment's sessions, add [`BASH_DEFAULT_TIMEOUT_MS` and `BASH_MAX_TIMEOUT_MS`](/docs/en/env-vars#variables) to its [environment variables](#set-environment-variables). Both take milliseconds. For example, `BASH_DEFAULT_TIMEOUT_MS=600000` makes 10 minutes the default.
376
366377## Setup scripts
367378
368379A setup script is a Bash script that runs when a new cloud session starts, before Claude Code launches. Use setup scripts to install dependencies, configure tools, or fetch anything the session needs that isn't pre-installed.
from line 394
383394A setup script has three constraints to write around:
384395
385396* **Exit zero**: if the script exits non-zero, the session fails to start. Append `|| true` to non-critical commands so an intermittent install failure doesn't block the session.
386* **Finish within five minutes**: keep the script's total runtime under roughly five minutes so the [environment cache](#environment-caching) can build. Run independent installs in parallel with `&` and `wait`, and move any single download that won't fit into a [SessionStart hook](#setup-scripts-vs-sessionstart-hooks) that launches it in the background.
397* **Finish within five minutes**: keep the script's total runtime under roughly five minutes so the [environment cache](#environment-caching) can build. When setup takes longer than that, the environment isn't cached. Run independent installs in parallel with `&` and `wait`, and move any single download that won't fit into a [SessionStart hook](#setup-scripts-vs-sessionstart-hooks) that launches it in the background. If new sessions stall or fail during setup, see [New sessions hang or time out during setup](/docs/en/web-quickstart#new-sessions-hang-or-time-out-during-setup).
387398* **Network access for installs**: package installs need to reach registries. The default **Trusted** level covers [common package registries](#default-allowed-domains) including npm, PyPI, RubyGems, and crates.io; with **None** network access, installs fail.
388399
389400### Environment caching
390401
391The setup script runs the first time you start a session in an environment. After it completes, Anthropic snapshots the filesystem and reuses that snapshot as the starting point for later sessions. New sessions start with your dependencies, tools, and Docker images already on disk, and skip the setup script step. This keeps startup fast even when the script installs large toolchains or pulls container images.
402The setup script runs the first time you start a session in an environment. When setup completes within [roughly five minutes](#script-requirements), Anthropic snapshots the filesystem and reuses that snapshot as the starting point for later sessions. New sessions start with your dependencies, tools, and Docker images already on disk, and skip the setup script step. This keeps startup fast even when the script installs large toolchains or pulls container images. If setup takes longer than roughly five minutes, the environment isn't cached.
392403
393404The cache is a filesystem snapshot, so it keeps what the setup script writes to disk and loses anything that was only running. Packages you install, Docker images you pull, and files you write all carry over. A database the script started, a `docker compose up` stack, or any other background process doesn't; start those per session by asking Claude or with a [SessionStart hook](#setup-scripts-vs-sessionstart-hooks).
394405
errors Changed · +12 / -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.
from line 23
2323| `Server error mid-response. The response above may be incomplete.` | [Server errors](#the-response-above-may-be-incomplete) |
2424| `Connection lost mid-response` / `Your computer went to sleep mid-response` / `The response stopped arriving` | [Server errors](#the-response-above-may-be-incomplete) |
2525| `Connection closed mid-response` / `Response stalled mid-stream` | [Server errors](#the-response-above-may-be-incomplete) |
26| `Part of the response never arrived` / `The response stream was malformed` | [Server errors](#the-response-above-may-be-incomplete) |
27| `API Error: Content block not found` / `API Error: Content block already closed` | [Server errors](#the-response-above-may-be-incomplete) |
2628| `Connection lost before a response was produced` / `Your computer went to sleep before a response was produced` / `The response stalled before a response was produced` | [Automatic retries](#automatic-retries) |
2729| `Connection closed while thinking` / `Response stalled while thinking` | [Automatic retries](#automatic-retries) |
2830| `Connection lost while your computer was asleep` | [Automatic retries](#automatic-retries) |
from line 460
458460API Error: Connection lost mid-response. The response above may be incomplete.
459461API Error: Your computer went to sleep mid-response. The response above may be incomplete.
460462API Error: The response stopped arriving. The response above may be incomplete.
463API Error: Part of the response never arrived. The response above may be incomplete.
464API Error: The response stream was malformed. The response above may be incomplete.
461465```
462466
463467* `Server error mid-response`: a mid-stream overloaded or 5xx server error. This variant requires Claude Code v2.1.199 or later; before then that case discarded the partial output and reported the whole turn as an error.
464* `Connection lost mid-response`: the connection dropped.
468* `Connection lost mid-response`: the connection dropped. You also see this variant when a proxy or gateway ends the response body cleanly before the response has finished.
465469* `Your computer went to sleep mid-response`: Claude Code detected that your computer went to sleep while the response was streaming. Once your computer wakes, Claude Code treats the connection as broken and stops reading from it.
470* `Part of the response never arrived`: a stream event was dropped between the API and Claude Code, so a later event referenced content that never arrived. Before v2.1.281, this case ended the turn with `API Error: Content block not found`.
471* `The response stream was malformed`: an event arrived for a content block that had already finished.
466472* `The response stopped arriving`: the connection stayed open but stopped delivering data, so the streaming idle watchdog aborted it. Before v2.1.222, Claude Code could also report this failure on [gateway](/docs/en/gateways) connections reached through `ANTHROPIC_BASE_URL` or `ANTHROPIC_AWS_BASE_URL` while the server's keep-alive pings were still arriving, because it counted only parsed response events there; upgrading stops those spurious timeouts on those routes. Gateways reached through a provider base URL such as `ANTHROPIC_BEDROCK_BASE_URL` aren't wrapped by the byte watchdog; see [Streaming idle watchdogs](/docs/en/network-config#streaming-idle-watchdogs).
467473
468474Before v2.1.227, `Connection lost mid-response` read `Connection closed mid-response` and `The response stopped arriving` read `Response stalled mid-stream`.
469475
476When a dropped or duplicated stream event arrives before Claude has started any text or tool call, you don't see this notice:
477
478* If Claude had completed only its thinking, Claude Code re-issues the request. When the re-issued streams break the same way, the turn ends with `Part of the response never arrived and no response was produced. Try again.` or `The response stream was malformed and no response was produced. Try again.`
479* If nothing had completed, Claude Code re-sends the request without streaming instead. If you turned that fallback off with [`CLAUDE_CODE_DISABLE_NONSTREAMING_FALLBACK`](/docs/en/env-vars), the turn ends with `API Error: Content block not found` for a dropped event or `API Error: Content block already closed` for a duplicated one.
480
470481In four cases, Claude Code handles the failure without showing this notice right away:
471482
472483* Earlier in the response, Claude Code either retries the failure or ends the turn with a different error. See [Automatic retries](#automatic-retries).
from line 1857
18461857
18471858Amazon Bedrock reports this condition as `Input is too long for requested model.`, which Claude Code handles the same way. Before v2.1.217, Claude Code didn't recognize the Bedrock wording, so auto-compact never triggered on it and `/compact` failed with the same error.
18481859
1849A [Claude apps gateway](/docs/en/claude-apps-gateway-config#upstream-error-messages) reports this condition as `capability_rejected: prompt_too_long` when a cloud upstream rejects the request in the provider's own error shape. Claude Code treats the token the same as `Prompt is too long`. Before v2.1.228, Claude Code didn't recognize the token, so auto-compact didn't trigger on it.
1850
1851When automatic compaction ran on this turn and failed on an underlying error, such as an unavailable model or an authentication failure, the message names that error after a separator:
1852
1853```text theme={null}
1854Prompt is too long · automatic compaction failed: <the underlying error>
1855```
1856
1857Resolve the named error first; `/compact` fails on the same error until you do. Before v2.1.229, a failed automatic compaction surfaced `Prompt is too long` without the cause.
1858
1859When automatic compaction runs on this error, it normally summarizes your oldest exchanges and keeps the newest. As a last resort, Claude Code summarizes differently:
1860
1861* When it can't summarize any whole exchange, Claude Code keeps your newest prompt word for word and summarizes everything before it.
1862* In that case, when the conversation doesn't end with your prompt, Claude Code summarizes the whole conversation instead.
1863
1864Claude Code skips this recovery when the content it would carry forward holds no model reply and less than about 1,000 tokens of your own text, such as a short retry sent after an oversized paste. Run `/clear` to start fresh. Before v2.1.269, compaction failed whenever it couldn't summarize a whole exchange, so a session in that state hit this error again on every turn.
1865
1866A single-exchange conversation has no earlier turns to summarize. When automatic compaction would have run on one, Claude Code skips the attempt and explains what fills the request instead. When the API doesn't report token counts in its error, the message reads:
1867
1868```text theme={null}
1869Prompt is too long · this conversation is a single exchange and cannot be compacted — the request size comes mostly from system prompt, tool definitions, or attachments.
1870```
1871
1872When the API reports token counts in its error, Claude Code compares th
1860A [C
model-config Changed · +11 / -7 lines
from line 622
622622
623623The effort scale is calibrated per model, so the same level name does not represent the same underlying value across models.
624624
625Opus 5.5 [defaults to `medium`](#adjust-effort-level), one level below Opus 5's default of `high`. In Anthropic's testing, Opus 5.5 at `medium` matches or exceeds Opus 5 at `high` on coding and knowledge-work evaluations. At a given level, Opus 5.5 tends to think more per turn than Opus 5. When you move from Opus 5 to Opus 5.5, start at `medium` rather than carrying over the level you used on Opus 5. To test levels against your own work, see [Calibrate effort](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/prompting-claude-opus-5-5#calibrate-effort) in the Opus 5.5 prompting guide.
626
625627#### Use ultrathink for one-off deep reasoning
626628
627629Include `ultrathink` anywhere in your prompt to request deeper reasoning on that turn without changing your session effort setting. Claude Code recognizes the keyword and adds an in-context instruction. The effort level sent to the API is unchanged. Claude Code passes other phrases such as "think", "think hard", and "think more" through as ordinary prompt text and doesn't recognize them as keywords.
from line 809
807809| `ANTHROPIC_DEFAULT_HAIKU_MODEL` | The model to use for `haiku`, or [background functionality](/docs/en/costs#background-token-usage) |
808810| `CLAUDE_CODE_SUBAGENT_MODEL` | The default model for [subagents](/docs/en/sub-agents#choose-a-model), [agent team](/docs/en/agent-teams#specify-teammates-and-models) teammates, and [workflow](/docs/en/workflows) agents that aren't assigned a model another way. Accepts an alias such as `haiku` or a full model name. A per-invocation model or a definition's `model` field, including `inherit`, takes precedence. To change that, set [`CLAUDE_CODE_SUBAGENT_MODEL_FORCE`](/docs/en/sub-agents#run-every-subagent-on-one-model) |
809811
812On third-party providers, [Customize pinned model display and capabilities](#customize-pinned-model-display-and-capabilities) describes what a pinned model's row in the `/model` picker shows.
813
810814Note: `ANTHROPIC_SMALL_FAST_MODEL` is deprecated in favor of
811815`ANTHROPIC_DEFAULT_HAIKU_MODEL`.
812816
from line 935
931935
932936Claude Code automatically uses [prompt caching](/docs/en/prompt-caching) to optimize performance and reduce costs. You can disable prompt caching globally or for specific model tiers:
933937
934| Environment variable | Description |
935| ------------------------------- | ------------------------------------------------------------------------------------------------- |
936| `DISABLE_PROMPT_CACHING` | Set to `1` to disable prompt caching for all models. Takes precedence over the per-model settings |
937| `DISABLE_PROMPT_CACHING_HAIKU` | Set to `1` to disable prompt caching for Haiku models only |
938| `DISABLE_PROMPT_CACHING_SONNET` | Set to `1` to disable prompt caching for Sonnet models only |
939| `DISABLE_PROMPT_CACHING_OPUS` | Set to `1` to disable prompt caching for Opus models only |
940| `DISABLE_PROMPT_CACHING_FABLE` | Set to `1` to disable prompt caching for Fable models only |
938| Environment variable | Description |
939| ------------------------------- | ------------------------------------------------------------------------------------------------------------- |
940| `DISABLE_PROMPT_CACHING` | Set to `1` to disable prompt caching for all models. Takes precedence over the per-model settings |
941| `DISABLE_PROMPT_CACHING_HAIKU` | Set to `1` to disable prompt caching for the [default Haiku model](/docs/en/prompt-caching#disable-prompt-caching) |
942| `DISABLE_PROMPT_CACHING_SONNET` | Set to `1` to disable prompt caching for Sonnet models only |
943| `DISABLE_PROMPT_CACHING_OPUS` | Set to `1` to disable prompt caching for Opus models only |
944| `DISABLE_PROMPT_CACHING_FABLE` | Set to `1` to disable prompt caching for Fable models only |
941945
942946To choose the cache TTL for the main conversation and for subagents separately, see [choose the TTL yourself](/docs/en/prompt-caching#choose-the-ttl-yourself). For what triggers a cache miss, see [How Claude Code uses prompt caching](/docs/en/prompt-caching).
943947
monitoring-usage Changed · +3 / -3 lines
from line 33
3333
3434To verify a setup that exports metrics, check your backend for the `claude_code.session.count` metric, which Claude Code emits when a session starts. To verify a logs-only setup, submit a prompt and check for the `claude_code.user_prompt` event.
3535
36If nothing arrives, run `claude --debug` and check the debug log. Claude Code reports failures from the exporters you configure as `[3P telemetry]` errors, where 3P means third-party. Lines prefixed `[Anthropic telemetry]` describe [Anthropic's separate operational telemetry](/docs/en/data-usage#telemetry-services) and don't indicate a problem with your setup.
36If nothing arrives, start Claude Code with `claude --debug-file <path>` and check the log it writes to that path. Claude Code reports failures from the exporters you configure as `[3P telemetry]` errors, where 3P means third-party. Lines prefixed `[Anthropic telemetry]` describe [Anthropic's separate operational telemetry](/docs/en/data-usage#telemetry-services) and don't indicate a problem with your setup.
3737
3838For full configuration options, see the [OpenTelemetry specification](https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/protocol/exporter.md#configuration-options).
3939
from line 62
6262
6363### How managed settings lock the OTLP destination
6464
65When you set an `OTEL_EXPORTER_OTLP_*` variable in managed settings, Claude Code removes conflicting developer-set variables at startup and logs a warning you can see with `claude --debug`. What it removes depends on which variable you set:
65When you set an `OTEL_EXPORTER_OTLP_*` variable in managed settings, Claude Code removes conflicting developer-set variables at startup and logs a warning in the debug log. What it removes depends on which variable you set:
6666
6767* **Endpoints**: when you set `OTEL_EXPORTER_OTLP_ENDPOINT`, Claude Code removes every developer-set per-signal endpoint. Developers can't point one signal at a different collector, so you don't need to also set the per-signal endpoint variables in managed settings.
6868* **Protocols**: when you set `OTEL_EXPORTER_OTLP_PROTOCOL`, Claude Code removes every developer-set per-signal protocol.
from line 1425
14251425}
14261426```
14271427
1428To confirm events arrive, submit a prompt in a session running under this configuration and check your SIEM for the `claude_code.user_prompt` event. If nothing arrives, run `claude --debug` and check the debug log for `[3P telemetry]` export errors.
1428To confirm events arrive, submit a prompt in a session running under this configuration and check your SIEM for the `claude_code.user_prompt` event. If nothing arrives, start Claude Code with `claude --debug-file <path>` and check that log for `[3P telemetry]` export errors.
14291429
14301430## Backend considerations
14311431
prompt-caching Changed · +29 / -14 lines
### Connecting or removing an MCP server ### Connecting or disconnecting an MCP server
from line 60
6060
6161## Actions that invalidate the cache
6262
63These actions cause the next request to miss part or all of the cache. You see a one-time slower, more expensive turn, after which the new prefix is cached. Most of them are avoidable mid-task once you know they have a cost. A model switch can feel free until you notice the slower turn that follows.
63These actions can cause the next request to miss part or all of the cache. You see a one-time slower, more expensive turn, after which the new prefix is cached. Most of them are avoidable mid-task once you know they have a cost. A model switch can feel free until you notice the slower turn that follows.
6464
6565* [Switching models](#switching-models)
6666* [Changing effort level](#changing-effort-level)
6767* [Turning on fast mode](#turning-on-fast-mode)
68* [Connecting or disconnecting an MCP server](#connecting-or-disconnecting-an-mcp-server)
68* [Connecting or removing an MCP server](#connecting-or-removing-an-mcp-server)
6969* [Enabling or disabling a plugin](#enabling-or-disabling-a-plugin)
7070* [Denying an entire tool](#denying-an-entire-tool)
7171* [Compacting the conversation](#compacting-the-conversation)
from line 102
102102
103103The cost applies once per conversation. After the first fast mode turn, Claude Code keeps sending the header and varies only the request's speed setting, which is not part of the cache key. Turning fast mode off, the [automatic fallback to standard speed](/docs/en/fast-mode#handle-rate-limits) after a rate limit, and turning it back on later all keep the cache. If you [run out of usage credits](/docs/en/fast-mode#handle-rate-limits) mid-session, Claude Code retries each rejected fast mode request at standard speed the same way, so this fallback also keeps the cache. `/clear` and `/compact` reset this, since they rebuild the cache at those points anyway.
104104
105### Connecting or disconnecting an MCP server
105### Connecting or removing an MCP server
106106
107107Tool definitions sit in the system prompt layer, so the cache invalidates when the set of tool definitions in the request changes between turns. Toggling the [advisor tool](/docs/en/advisor) is an exception: its definition sits after the cache breakpoint, so enabling or disabling `/advisor` keeps the cached prefix intact. Whether an [MCP server](/docs/en/mcp) change does this depends on whether its tools are deferred by [tool search](/docs/en/mcp#scale-with-mcp-tool-search) or loaded into the prefix:
108108
109109* **Deferred tools**, the default on supported models: a server connecting, disconnecting, or changing its tool list only appends new content and doesn't disturb anything already cached.
110* **Tools loaded into the prefix**: any change to them invalidates the cache. This happens when [tool search is unavailable or disabled](/docs/en/mcp#configure-tool-search), such as on Google Cloud's Agent Platform models earlier than the Claude 4.5 generation, with a custom `ANTHROPIC_BASE_URL` gateway, or on 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 that the deployment rejects tool search. It also happens for a server or tool marked [`alwaysLoad`](/docs/en/mcp#exempt-a-server-from-deferral), and for definitions kept upfront by [threshold-based loading](/docs/en/mcp#configure-tool-search).
110* **Tools loaded into the prefix**: adding a definition invalidates the cache, and so does removing one on purpose. This is the case when [tool search is unavailable or disabled](/docs/en/mcp#configure-tool-search), such as on Google Cloud's Agent Platform models earlier than the Claude 4.5 generation, with a custom `ANTHROPIC_BASE_URL` gateway, or on 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 that the deployment rejects tool search.
111111
112When tools load into the prefix, the most common cause of an invalidation is a server connecting or disconnecting mid-session, which can happen without any action on your part: a stdio server's process exits, an HTTP session expires, or a server [reconnects automatically after a transient failure](/docs/en/mcp#automatic-reconnection). A connected server can also push a [dynamic tool update](/docs/en/mcp#dynamic-tool-updates) that changes its tool list.
112Without tool search, whether a mid-session server change invalidates the cache depends on what changed. For each change, this table gives whether the cache is kept and what happens to the tool definitions in the next request.
113113
114| Mid-session change | Cache | Tool definitions in the next request |
115| ---------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
116| A server connects, or a [dynamic tool update](/docs/en/mcp#dynamic-tool-updates) adds tools | Invalidated | The new definitions are added |
117| A server drops out with no action on your part, such as a stdio server's process exiting | Kept | The server's definitions stay unchanged. A call to one of its tools returns an error instead of running |
118| A remote server [reconnects automatically](/docs/en/mcp#automatic-reconnection) after its connection drops | Kept, unless a request sent while the server reconnects adds the `WaitForMcpServers` tool, which invalidates the cache once | The server's definitions stay unchanged. A request sent while the server reconnects can add `WaitForMcpServers` when the conversation hasn't listed it yet, and the tool then stays listed for the rest of the conversation |
119| You remove a tool on purpose, such as with a [deny rule](#denying-an-entire-tool) or by disabling its server in `/mcp` | Invalidated | The definition is removed |
120
121When you resume a conversation whose tools load into the prefix, one of its MCP servers can still be connecting as the first request goes out. If the transcript recorded that server's tool definitions, that request includes them as recorded, so it doesn't change when the server finishes connecting with the same tools.
122
114123Editing your MCP config does not by itself change the cache. The new config takes effect only after a restart, which is when the server connects or disconnects.
115124
116125### Enabling or disabling a plugin
from line 132
123132
124133#### Plugins that provide MCP servers
125134
126When you enable or disable a plugin that provides [MCP servers](/docs/en/plugins/components#mcp-servers), Claude Code follows the same rules as when you [connect or disconnect an MCP server](#connecting-or-disconnecting-an-mcp-server):
135When you enable or disable a plugin that provides [MCP servers](/docs/en/plugins/components#mcp-servers), Claude Code follows the same rules as when you [connect or remove an MCP server](#connecting-or-removing-an-mcp-server):
127136
128137* If Claude Code defers the server's tools, it keeps the cache.
129138* If Claude Code loads them into the prefix, the next request re-reads the entire conversation.
from line 341
332341
333342Disabling caching is occasionally useful when debugging caching behavior with a specific model or provider. To turn it off, set one of these environment variables to `1`:
334343
335| Variable | Effect |
336| ------------------------------- | ----------------------- |
337| `DISABLE_PROMPT_CACHING` | Disable for all models |
338| `DISABLE_PROMPT_CACHING_HAIKU` | Disable for Haiku only |
339| `DISABLE_PROMPT_CACHING_SONNET` | Disable for Sonnet only |
340| `DISABLE_PROMPT_CACHING_OPUS` | Disable for Opus only |
341| `DISABLE_PROMPT_CACHING_FABLE` | Disable for Fable only |
344| Variable | Effect |
345| ------------------------------- | ----------------------------------- |
346| `DISABLE_PROMPT_CACHING` | Disable for all models |
347| `DISABLE_PROMPT_CACHING_HAIKU` | Disable for the default Haiku model |
348| `DISABLE_PROMPT_CACHING_SONNET` | Disable for Sonnet only |
349| `DISABLE_PROMPT_CACHING_OPUS` | Disable for Opus only |
350| `DISABLE_PROMPT_CACHING_FABLE` | Disable for Fable only |
351
352`DISABLE_PROMPT_CACHING_HAIKU` applies to the default Haiku model, the model the `haiku` alias resolves to. It disables caching wherever that model runs, including the main conversation when it is your main model. Covering the main conversation requires Claude Code v2.1.283 or later.
353
354The variable also covers a background model you set with the deprecated `ANTHROPIC_SMALL_FAST_MODEL` variable, when that model differs from your main model.
355
356A different Haiku version that you pin as your main model keeps caching; set `DISABLE_PROMPT_CACHING` to disable caching for it.
342357
343358To set caching policy across an organization, put any of these or the [TTL variables](#cache-lifetime) in the `env` block of [managed settings](/docs/en/managed-settings). For normal use, leave caching enabled.
344359
skills Changed · +4 / -2 lines
from line 266
266266 Edit a skill during a session
267267</h3>
268268
269Claude Code watches skill directories for file changes, except in [bare mode](/docs/en/headless#start-faster-with-bare-mode). When you add, edit, or remove a skill under `~/.claude/skills/`, the project `.claude/skills/`, or a `.claude/skills/` inside an `--add-dir` directory, Claude Code picks up the change within the current session, without a restart. If you create a top-level skills directory that didn't exist when the session started, restart Claude Code so it can watch the new directory.
269Claude Code watches skill directories for file changes, except in [bare mode](/docs/en/headless#start-faster-with-bare-mode). When you add, edit, or remove a skill under `~/.claude/skills/`, the project `.claude/skills/`, or a `.claude/skills/` inside an `--add-dir` directory, Claude Code picks up the change within the current session, without a restart.
270270
271If you create a top-level skills directory that didn't exist when the session started, run [`/reload-skills`](/docs/en/commands#all-commands) to pick up the skills you put there. Claude Code isn't watching that directory yet, so run `/reload-skills` again after each later change there.
272
271273Live change detection covers `SKILL.md` text only. For a skill folder that is also a [plugin](/docs/en/plugins/loading#plugins-shared-through-a-repository), changes to `hooks/`, `.mcp.json`, `agents/`, and `output-styles/` need `/reload-plugins` to take effect.
272274
273275### Remove a skill
from line 1095
10931095
10941096### Skill descriptions are cut short
10951097
1096Claude Code loads a listing of skill names and descriptions into context so Claude knows what's available. The listing always contains every skill name, but if you have many skills, Claude Code shortens descriptions to fit the listing's character budget, which can strip the keywords Claude needs to match your request. The budget scales at 1% of the model's context window. When the listing overflows, Claude Code drops descriptions starting with the skills you invoke least, so the skills you use most keep their full text.
1098Claude Code loads a listing of skill names and descriptions into context so Claude knows what's available. The listing always contains every skill name, but if you have many skills, Claude Code drops some descriptions to fit the listing's character budget, which removes the keywords Claude needs to match your request. The budget scales at 1% of the model's context window. When the listing overflows, Claude Code drops descriptions starting with the skills you invoke least, so the skills you use most keep their full text.
10971099
10981100Run `/doctor` for an estimate of the listing's context cost and its biggest contributors. To find skills worth turning off, run [`/skill-doctor`](#find-unused-skills). When the listing exceeds its budget, Claude Code also writes a warning to the debug log, visible with [`--debug`](/docs/en/cli-reference#cli-flags).
10991101
worktrees Changed · +3 / -3 lines
from line 50
5050When you exit an interactive worktree session, Claude checks the worktree for work that removal would delete: changed or untracked files, uncommitted work inside checked-out submodules, and new commits.
5151
5252* **The worktree is clean**: for an unnamed session, Claude removes the worktree and its branch automatically. A [named](/docs/en/sessions#name-your-sessions) session prompts you first so you can keep the worktree for later
53* **The worktree has work in it**: Claude prompts you to keep or remove the worktree. Keeping preserves the directory and branch so you can return later. Removing deletes the worktree directory and its branch, along with all the work in them
53* **The worktree has work in it**: Claude prompts you to keep or remove the worktree. Keeping preserves the directory and branch. To return later, run the `claude --worktree <name> --resume` command that Claude Code prints on exit. Removing deletes the worktree directory and its branch, along with all the work in them
5454* **The worktree's state can't be verified**: when Claude Code can't count the worktree's changes or can't inspect its submodule checkouts, it prompts you rather than removing the worktree automatically. The prompt names what it couldn't check
5555
5656Non-interactive runs with `-p` have no exit prompt, so Claude doesn't clean up their worktrees, and Claude Code leaves the lock it took on each one at creation in place until a later session's [stale-lock sweep](#clean-up-subagent-and-background-session-worktrees) releases it. To remove one, run `git worktree remove`; if git refuses because the worktree is locked, run `git worktree unlock` on it first.
from line 59
5959
6060## Resume a worktree session
6161
62When you resume a session that was inside a worktree, Claude Code returns the session to that worktree. This holds for interactive resumes, for `--continue` and `--resume` in [non-interactive mode](/docs/en/headless) with `-p`, and for the Agent SDK. Back inside the worktree, Claude can still exit it with the [`ExitWorktree`](/docs/en/tools-reference) tool.
62When you resume a session that ended inside a worktree without [exiting it](#clean-up-worktrees), Claude Code returns the session to that worktree. This holds for interactive resumes, for `--continue` and `--resume` in [non-interactive mode](/docs/en/headless) with `-p`, and for the Agent SDK. `--continue` picks the most recent session recorded under the directory you launch from. Back inside the worktree, Claude can still exit it with the [`ExitWorktree`](/docs/en/tools-reference) tool.
6363
6464Before returning the session to its worktree, Claude Code verifies that the worktree is still a checkout separate from the main one, and declines to re-enter a worktree that fails the check. For a git worktree, the check reads its git metadata. A worktree without git metadata, such as one a [`WorktreeCreate` hook](#non-git-version-control) created, can pass the check; the cases Claude Code still refuses are listed with their recoveries under [Claude Code refuses to use a worktree](#claude-code-refuses-to-use-a-worktree). For the messages and how to recover from each, see [The session resumes outside its worktree](#the-session-resumes-outside-its-worktree).
6565
6666Where you launch from, and how you resume, change what Claude Code re-enters:
6767
68* **Launch directory**: resume from the main checkout or another directory of the repository. Claude Code re-enters a worktree it created with git under `.claude/worktrees/` even when you launch from inside it. When you launch from inside any other worktree, Claude Code re-enters it only if it can vouch for it from there: a worktree that is its own repository, one without git metadata, or a launch from a subdirectory of a worktree you created with `git worktree add` declines, so launch those from the main checkout.
68* **Launch directory**: resume with `--resume` from the main checkout or another directory of the repository. Claude Code re-enters a worktree it created with git under `.claude/worktrees/` even when you launch from inside it. When you launch from inside any other worktree, Claude Code re-enters it only if it can vouch for it from there: a worktree that is its own repository, one without git metadata, or a launch from a subdirectory of a worktree you created with `git worktree add` declines, so launch those from the main checkout.
6969* **`--fork-session`**: the forked session starts in the directory you launched Claude from, and Claude Code leaves the original session's worktree untouched.
7070* **Deleted worktree**: if the worktree directory no longer exists, Claude Code resumes the session in the directory you launched Claude from. It tells you the worktree is gone and clears the session's worktree binding.
7171
best-practices Changed · +1 / -1 lines
from line 40
4040
4141* **In one prompt**: ask Claude to run the check and iterate in the same message, as in the table above.
4242* **Across a session**: set the check as a [`/goal` condition](/docs/en/goal). A separate evaluator re-checks it after every turn and Claude keeps working until the goal resolves. If Claude stalls, Claude Code eventually stops the run with the goal still set — see [how /goal evaluation works](/docs/en/goal#how-evaluation-works).
43* **As a deterministic gate**: a [Stop hook](/docs/en/hooks#stop) runs your check as a script and blocks the turn from ending until it passes. Claude Code overrides the hook and ends the turn after 8 consecutive blocks.
43* **As a deterministic gate**: a [Stop hook](/docs/en/hooks#stop) runs your check as a script and blocks the turn from ending until it passes. [Stop input](/docs/en/hooks#stop-input) covers the cap on consecutive blocks.
4444* **By a second opinion**: a [verification subagent](/docs/en/sub-agents) or a [dynamic workflow](/docs/en/workflows) that checks its own findings has a fresh model try to refute the result, so the agent doing the work isn't the one grading it.
4545
4646Each step trades setup for attention. The prompt version works on any task today. The `/goal` and Stop hook versions are what let an unattended run finish correctly without you.
channels Changed · +1 / -1 lines
from line 343
343343
344344During the preview, `--channels` only accepts plugins from an Anthropic-maintained allowlist, or from your organization's allowlist if an admin has set [`allowedChannelPlugins`](#restrict-which-channel-plugins-can-run). The channel plugins in [claude-plugins-official](https://github.com/anthropics/claude-plugins-official/tree/main/external_plugins) are the default approved set. If you pass something that isn't on the effective allowlist, Claude Code starts normally but the channel doesn't register, and the startup notice tells you why.
345345
346To test a channel you're building, use `--dangerously-load-development-channels`. See [Test during the research preview](/docs/en/channels-reference#test-during-the-research-preview) for information about testing custom channels that you build.
346To test a channel you're building, pass it to `--dangerously-load-development-channels` in the `plugin:<name>@<marketplace>` or `server:<name>` form. See [Test during the research preview](/docs/en/channels-reference#test-during-the-research-preview) for information about testing custom channels that you build.
347347
348348Report issues or feedback on the [Claude Code GitHub repository](https://github.com/anthropics/claude-code/issues).
349349
claude-code-on-the-web Changed · +1 / -0 lines
from line 365
365365Before relying on cloud sessions for a workflow, account for these constraints:
366366
367367* **Rate limits**: cloud sessions share rate limits with all other Claude and Claude Code usage within your account. Running multiple tasks in parallel consumes more rate limits proportionately. There is no separate compute charge for the cloud VM.
368* **Time limits**: commands Claude runs and SessionStart hooks have default timeouts you can change, and a setup script is cached only when it finishes in roughly five minutes. See [Time limits](/docs/en/cloud-environments#time-limits)
368369* **Repository authentication**: you can only pull a cloud session into your terminal when you are authenticated to the same account
369370* **Platform restrictions**: repository cloning and pull request creation require GitHub. Self-hosted [GitHub Enterprise Server](/docs/en/github-enterprise-server) instances are supported for Team and Enterprise plans. You can send a GitLab, Bitbucket, or other non-GitHub repository to a cloud session as a [local bundle](#send-local-repositories-without-github) by setting `CCR_FORCE_BUNDLE=1`, but the session can't push results back to that remote
370371* **Organization IP allowlist**: cloud sessions call the Anthropic API from Anthropic-managed infrastructure, not your network, while sessions in a [self-hosted environment](/docs/en/self-hosted-environments) call it from your own network. If your organization has [IP allowlisting](https://support.claude.com/en/articles/13200993-restrict-access-to-claude-with-ip-allowlisting) enabled, every Anthropic-hosted cloud session fails with an authentication error. The same applies to [Code Review](/docs/en/code-review) and to [routines](/docs/en/routines) that run on Anthropic-hosted environments; a routine routed to a self-hosted environment calls the API from your own network. Contact [Anthropic support](https://support.claude.com/) to exempt Anthropic-hosted services from your organization's IP allowlist.
env-vars 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.
github-enterprise-server Changed · +1 / -1 lines
from line 212
212212
213213### Session start fails with `Unable to get organization UUID`
214214
215Cloud sessions require a Team or Enterprise organization. Sign in with `/login` using your organization account. If you authenticate with an API key instead, cloud sessions fail earlier with a message asking you to run `/login`.
215Sign in with `/login` using your organization account. If you authenticate with an API key instead, cloud sessions fail earlier with a message asking you to run `/login`.
216216
217217## Related resources
218218
goal Changed · +1 / -1 lines
from line 129
129129
130130If a turn fails on an error that won't clear until you fix it, Claude Code clears the goal and prints a warning naming the cause. The warning starts with `Goal cleared after an unrecoverable error` and ends with `Run /goal again to continue`. Fix the cause, then [set the goal again](#set-a-goal) with `/goal <condition>`. Four kinds of failure clear the goal:
131131
132* An authentication failure, when Claude Code manages its own credentials. When a host manages them for you, such as the desktop app, the VS Code extension, or a [cloud session](/docs/en/claude-code-on-the-web), Claude Code leaves the goal active because the host restores access on its own.
132* An authentication failure, when Claude Code manages its own credentials. When a host manages them for you, such as the desktop app or a [cloud session](/docs/en/claude-code-on-the-web), Claude Code leaves the goal active because the host restores access on its own.
133133* An exhausted credit balance
134134* A context overflow that [auto-compaction](/docs/en/model-config#set-the-auto-compact-window) couldn't clear
135135* A model that isn't available
hooks 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.
from line 2545
25452545
25462546#### Stop input
25472547
2548In addition to the [common input fields](#common-input-fields), Stop hooks receive `stop_hook_active`, `last_assistant_message`, `background_tasks`, and `session_crons`. The `stop_hook_active` field is `true` when Claude Code is already continuing as a result of a stop hook. Check this value or process the transcript to avoid blocking on a condition that will never resolve. Claude Code overrides the hook and ends the turn after 8 consecutive blocks.
2548In addition to the [common input fields](#common-input-fields), Stop hooks receive `stop_hook_active`, `last_assistant_message`, `background_tasks`, and `session_crons`. The `stop_hook_active` field is `true` when Claude Code is already continuing as a result of a stop hook. Check this value or process the transcript to avoid blocking on a condition that will never resolve. Claude Code applies an 8-consecutive-continuation cap: after stop hooks have continued the turn eight times in a row, Claude Code overrides the next block and ends the turn. To raise the cap, set [`CLAUDE_CODE_STOP_HOOK_BLOCK_CAP`](/docs/en/env-vars).
25492549
25502550The `last_assistant_message` field contains the text content of Claude's final response, so hooks can access it without parsing the transcript file. For hooks that act on the just-completed turn, such as read-aloud or notification hooks, use this field rather than reading `transcript_path`: the transcript file isn't guaranteed to include the final message at Stop time on all versions.
25512551
from line 2655
26552655 "transcript_path": "/Users/.../.claude/projects/.../00893aaf-19fa-41d2-8238-13269b9b3ca0.jsonl",
26562656 "cwd": "/Users/...",
26572657 "hook_event_name": "StopFailure",
2658 "error": "rate_limit",
2659 "error_details": "429 Too Many Requests",
2660 "last_assistant_message": "API Error: Rate limit reached"
2661}
2662```
2663
2664StopFailure hooks have no decision contr
2658
llm-gateway-connect Changed · +1 / -1 lines
from line 122
122122<Tabs>
123123 <Tab title="Bash or Zsh">
124124 ```bash theme={null}
125 curl -X POST "$ANTHROPIC_BASE_URL/v1/messages" \
125 curl -sS -w '\n%{http_code}\n' -X POST "$ANTHROPIC_BASE_URL/v1/messages" \
126126 -H "Authorization: Bearer $ANTHROPIC_AUTH_TOKEN" \
127127 -H "anthropic-version: 2023-06-01" \
128128 -H "content-type: application/json" \
llm-gateway-protocol Changed · +4 / -0 lines
from line 61
6161
6262Stream inference responses. Claude Code reads the stream as it arrives, so if your gateway buffers complete responses before relaying them, Claude Code stalls.
6363
64Deliver each response's full event sequence without dropping, duplicating, or reordering events. When an event references a content block whose `content_block_start` never arrived, or a block whose `content_block_stop` already arrived, Claude Code stops reading the stream at that event instead of applying it, so a duplicated `content_block_stop` can't run the same tool call twice. [The response above may be incomplete](/docs/en/errors#the-response-above-may-be-incomplete) describes what the user sees, under the `Part of the response never arrived` and `The response stream was malformed` variants.
65
66Relay each response through its final `message_delta` and `message_stop` events before ending the body. A body that ends after a `message_delta` carrying a `stop_reason`, with no content block still open and no content block event after that frame, counts as complete even when `message_stop` is missing. A body that your gateway ends cleanly any earlier, once a content block has started, is treated the same as a dropped connection: [Automatic retries](/docs/en/errors#automatic-retries) says when Claude Code re-issues the request, and [The response above may be incomplete](/docs/en/errors#the-response-above-may-be-incomplete) covers what it keeps once visible content has arrived. Claude Code keeps the `stop_reason` a `message_delta` delivers, so a later usage-only `message_delta` whose `delta` has `stop_reason: null` or no `stop_reason` key doesn't clear it.
67
6468When the client speaks the Amazon Bedrock format, relay the `InvokeModelWithResponseStream` response body and its `Content-Type: application/vnd.amazon.eventstream` header unmodified, and don't convert the stream to server-sent events. See [Streaming errors behind a gateway or proxy](/docs/en/amazon-bedrock#streaming-errors-behind-a-gateway-or-proxy).
6569
6670Forward keep-alive pings as well. On connections through `ANTHROPIC_BASE_URL` or `ANTHROPIC_AWS_BASE_URL`, Claude Code counts every byte your gateway relays, including SSE `ping` events and comment lines, and aborts a stream that goes silent for 300 seconds by default. The upstream's pings are the only traffic during long thinking pauses, so if your gateway strips or buffers them, Claude Code aborts the stream during those pauses; [Automatic retries](/docs/en/errors#automatic-retries) covers what an aborted stream reports based on how far the response had progressed. An upstream that sends no pings at all, such as Amazon Bedrock's binary event-stream, leaves those pauses with nothing to forward. When translating from such an upstream, emit your own `ping` events during silent gaps. Gateways reached through `ANTHROPIC_BEDROCK_BASE_URL`, `ANTHROPIC_VERTEX_BASE_URL`, or `ANTHROPIC_FOUNDRY_BASE_URL` aren't wrapped by this byte-level watchdog, even when they relay the Anthropic Messages format; there, a [5-minute idle timeout](/docs/en/env-vars) aborts a silent stream instead, and on `ANTHROPIC_BEDROCK_BASE_URL` connections you can add the byte watchdog with [`CLAUDE_ENABLE_BYTE_WATCHDOG_BEDROCK`](/docs/en/env-vars).
memory Changed · +4 / -0 lines
from line 243
243243ln -s ~/company-standards/security.md .claude/rules/security.md
244244```
245245
246If you point a `.claude/rules/` or `CLAUDE.md` symlink at a network path such as the UNC share `\\server\share` or a path under `/net` or `/Network`, the linked instructions don't load. Claude Code doesn't follow the link, because looking up such a path can contact the host it names. `\\wsl$` paths don't count as network paths.
247
246248#### User-level rules
247249
248250Personal rules in `~/.claude/rules/` apply to every project on your machine. Use them for preferences that aren't project-specific:
from line 582
580582### My CLAUDE.md is too large
581583
582584Files over 200 lines consume more context and may reduce adherence. Claude Code skips a file over 4 MiB. Use [path-scoped rules](#path-specific-rules) to load instructions only when Claude works with matching files, or trim content that isn't needed in every session. Splitting into [`@path` imports](#import-additional-files) helps organization but doesn't reduce context, since imported files load at launch.
585
586If one of your instruction files is over the recommended length, you see a warning at startup and when you run `/status`. You also see a warning when files that are each within that length add up past a combined limit at session start. Each CLAUDE.md, rules file, and `@path` import counts as a separate file.
583587
584588The [`/doctor`](/docs/en/commands#all-commands) checkup proposes trims for a checked-in CLAUDE.md: it cuts content Claude can derive from the codebase, such as directory layouts, dependency lists, and architecture overviews, and keeps pitfalls, rationale, and conventions that differ from tool defaults. The trim check requires Claude Code v2.1.206 or later.
585589
sandbox-environments Changed · +1 / -1 lines
from line 101
101101On Linux and WSL2, the runtime applies write grants only to paths that already exist. In a fresh environment, create Claude Code's configuration paths before the first launch:
102102
103103```bash theme={null}
104mkdir -p ~/.claude && echo '{}' > ~/.claude.json
104mkdir -p ~/.claude && { [ -f ~/.claude.json ] || echo '{}' > ~/.claude.json; }
105105```
106106
107107Once the settings file is in place, launch Claude Code with `npx` and pass `claude` as the command to wrap:
statusline Changed · +2 / -2 lines
from line 159
159159
160160Claude Code captures your script's output instead of connecting it directly to the terminal, so `tput cols` and language-level width detection cannot read the terminal size from inside the script. Read the `COLUMNS` and `LINES` environment variables instead. Claude Code sets these to the current terminal dimensions before running your script.
161161
162<Note>The status line runs locally and does not consume API tokens. It temporarily hides during certain UI interactions, including autocomplete suggestions, the help menu, and permission prompts.</Note>
162<Note>The status line runs locally and does not consume API tokens. It temporarily hides during certain UI interactions, including the help menu and permission prompts.</Note>
163163
164164## Available data
165165
from line 1120
11201120* On Windows with Git Bash installed, backslashes in the `command` path are likely being consumed as escape characters before the script runs. Use forward slashes in the path. See [Windows configuration](#windows-configuration).
11211121* If `disableAllHooks` is `true` outside managed settings after [settings precedence](/docs/en/hooks#disable-or-remove-hooks) applies, Claude Code runs only a `statusLine` from managed settings, and with no managed `statusLine` the status line is disabled. Remove the setting, or set it to `false` in the file that sets it, to re-enable. See [`disableAllHooks`](/docs/en/settings-reference#disableallhooks).
11221122* If your organization sets `allowManagedHooksOnly` in managed settings, your custom status line disappears without warning: you can only get a status line from a `statusLine` value in those managed settings. See [what runs under `allowManagedHooksOnly`](/docs/en/settings-reference#what-runs-under-allowmanagedhooksonly) for the full behavior, and ask your administrator whether this setting applies to you.
1123* Run `claude --debug` to log the exit code and stderr from the first status line invocation in a session
1123* Run `claude --debug` to log your script's stderr on every status line invocation, and its exit code on the first invocation in a session
11241124* Ask Claude to read your settings file and execute the `statusLine` command directly to surface errors
11251125
11261126**Status line shows `--` or empty values**
workflows Changed · +3 / -1 lines
from line 456
456456
457457To turn workflows off for your whole organization, set `"disableWorkflows": true` in [managed settings](/docs/en/server-managed-settings), or use the toggle on the [Claude Code admin settings](https://claude.ai/admin-settings/claude-code) page.
458458
459When workflows are disabled, the bundled workflow commands and the `/workflow-authoring` skill are unavailable, the `ultracode` keyword no longer triggers a run, and `ultracode` is removed from the `/effort` menu.
459When workflows are disabled, the bundled workflow commands and the `/workflow-authoring` skill are unavailable, the `ultracode` keyword no longer triggers a run, and `ultracode` is removed from the `/effort` menu. A run that was already in progress keeps going.
460
461No setting turns off [ultracode](#let-claude-decide-with-ultracode) alone. To keep workflows but rule out ultracode, set an [effort cap](/docs/en/model-config#organization-effort-limits) below `xhigh` on the models you want to cover. Ultracode needs `xhigh`, so it's then [unavailable](/docs/en/model-config#when-ultracode-is-available) on those models.
460462
461463## Related resources
462464