What
- When 1M-context usage requires usage credits, the guidance message now checks whether the session is non-interactive or self-hosted-like, and whether the extra-usage-credits command is actually available (which itself depends on the
DISABLE_EXTRA_USAGE_COMMANDsetting). Depending on the result, the message either offers the/usage-creditscommand or only mentions/model. - Similarly, when a requested model isn't available, the suggestion of switching via
/modelversus the--modelflag now takes into account whether the session is interactive, and whether it's running in contexts like the SDK, CLI, orclaude-code-github-action.
Why
Previously these suggestions could point users toward commands or flags that don't actually work in their current context (for example, suggesting an interactive /model command inside a non-interactive script or GitHub Action). The messages now match what's actually usable in the situation.
/model
Use Claude Code in the cloud modified, high confidence
Auto-compaction runs automatically when the context window approaches capacity. Cloud sessions set [`CLAUDE_AUTOCOMPACT_PCT_OVERRIDE`](/docs/en/env-vars) themselves, so compaction triggers partway through the [auto-compact window](/docs/en…see the edit
/model
SEP-2640: Skills Extension modified, high confidence
* **Fragmented distribution.** A server and the skill that teaches an agent to use it are versioned, discovered, and installed separately. Users installing a server from a registry have no signal that a companion skill exists. ([problem st…see the edit
/model
Configure server-managed settings modified, medium confidence
If your devices are enrolled in an MDM or endpoint management solution, endpoint-managed settings provide stronger security guarantees because the settings file can be protected from user modification at the OS level. Endpoint-managed sett…see the edit
/model
sent_count = len(messages sent in the compaction request) modified, medium confidence
Send the `compact-2026-09-04` beta header on the request that asks for the summary and on every later request that carries the signed block. On-demand compaction is available on the Claude API but not on Amazon Bedrock or Google Cloud. It …see the edit
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Added since
A small documentation edit on On a remote host without a browser: touched a line naming --model after this was published.
--model claude-opus-5-5 \cli-sdks-libraries/cli/authentication see the edit
Confirmed since
Anthropic's documentation has since written up --model, on Model configuration.
The plan-mode Opus phase uses the same context window as the `opus` model setting, and the execution phase uses the same window as `sonnet`. When `opus` and `sonnet` resolve to models that run with the [1M context window](#extended-context…model-config see the edit
Added since
A small documentation edit on Claude Code gateway compatibility guide touched a line naming /model after this was published.
The request is `GET /v1/models?limit=1000` with a timeout of 3 seconds by default, and any redirect is treated as failure so the credential can't leak to a redirect target. A gateway that responds slower than the timeout, or one that redir…llm-gateway-protocol see the edit
Confirmed since
Anthropic's documentation has since written up /model, on Environment variables.
| `CLAUDE_CODE_GATEWAY_MODEL_DISCOVERY_TIMEOUT_MS` | Timeout in milliseconds for the [gateway model discovery](/docs/en/llm-gateway-protocol#model-discovery) request that `CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY` turns on (default: `300…env-vars see the edit
Two sources agreeTwo things we can check say the same as this entry.
Anthropic's documentation agrees
/model on Use Claude Code in the cloud
Anthropic's release notes agree
Fixed claude agents losing --model, --effort, --permission-mode, --allow-dangerously-skip-permissions and --agent after an auto-update…