An entry is only as true as the day it was written.
Flags move on the server after a release ships. Anthropic's own notes land days later and their documentation later still. Sometimes an entry turns out to be wrong and gets corrected by hand. All of that is recorded against the entry it happened to, and this is the whole of it in the order it happened.
Nothing here replaces an entry. Every line links to the page its release published, still saying what that release said, because a correction is added and dated rather than swapped in.
| When | Entry | What happened |
|---|---|---|
v2.1.267CLAUDE_CODE_EXTRA_BODY |
Documentation arrivedAnthropic's documentation has since written up CLAUDE_CODE_EXTRA_BODY, on Claude Code changelog.
the line in the docs* Changed `/bug` and `/feedback` reports to include only model-behavior params (model, system prompt, tools) from the last API request, omitting request metadata and `CLAUDE_CODE_EXTRA_BODY` fields |
|
v2.1.271/feedback |
Documentation arrivedAnthropic's documentation has since written up /feedback, on Claude Code changelog.
the line in the docs* Changed `/bug` and `/feedback` reports to include only model-behavior params (model, system prompt, tools) from the last API request, omitting request metadata and `CLAUDE_CODE_EXTRA_BODY` fields |
|
v2.1.271CLAUDE_CODE_AUTO_MODE_SERVER |
Documentation arrivedAnthropic's documentation has since written up CLAUDE_CODE_AUTO_MODE_SERVER, on Claude Code changelog.
the line in the docs* Changed auto mode on Bedrock, Vertex and Foundry to use the local classifier by default for now; set `CLAUDE_CODE_AUTO_MODE_SERVER=1` to use the platform's server-side classifier |
|
v2.1.248NODE_EXTRA_CA_CERTS |
Documentation arrivedAnthropic's documentation has since written up NODE_EXTRA_CA_CERTS, on Claude Code changelog.
the line in the docs* Improved the in-session SSL certificate and proxy connection errors to name the error code and what to fix, such as `NODE_EXTRA_CA_CERTS` for an untrusted corporate CA |
|
v2.1.268/web-setup |
Documentation arrivedAnthropic's documentation has since written up /web-setup, on Claude Code changelog.
the line in the docs* Improved `/web-setup` errors: a refused GitHub token now lists the likely reasons and the fix, and a connection failure names a configured proxy or TLS certificate problem |
|
v2.1.268/web-setup |
Documentation arrivedAnthropic's documentation has since written up /web-setup, on Claude Code changelog.
the line in the docs* Improved `/web-setup` errors: a refused GitHub token now lists the likely reasons and the fix, and a connection failure names a configured proxy or TLS certificate problem |
|
v2.1.268/web-setup |
Documentation arrivedAnthropic's documentation has since written up /web-setup, on Claude Code changelog.
the line in the docs* Improved `/web-setup` errors: a refused GitHub token now lists the likely reasons and the fix, and a connection failure names a configured proxy or TLS certificate problem |
|
v2.1.268/web-setup |
Documentation arrivedAnthropic's documentation has since written up /web-setup, on Claude Code changelog.
the line in the docs* Improved `/web-setup` errors: a refused GitHub token now lists the likely reasons and the fix, and a connection failure names a configured proxy or TLS certificate problem |
|
v2.1.259/install-github-app |
Documentation arrivedAnthropic's documentation has since written up /install-github-app, on Claude Code changelog.
the line in the docs* Fixed `/install-github-app` reporting a SAML single sign-on block as "admin permissions required" |
|
v2.1.234tui |
Documentation arrivedAnthropic's documentation has since written up tui, on Claude Code changelog.
the line in the docs* Fixed `/tui` refusing to restart because of an agent-team teammate that had already finished its work and was no longer shown in the agents panel |
|
v2.1.248/tui |
Documentation arrivedAnthropic's documentation has since written up /tui, on Claude Code changelog.
the line in the docs* Fixed `/tui` refusing to restart because of an agent-team teammate that had already finished its work and was no longer shown in the agents panel |
|
v2.1.234/tui |
Documentation arrivedAnthropic's documentation has since written up /tui, on Claude Code changelog.
the line in the docs* Fixed `/tui` refusing to restart because of an agent-team teammate that had already finished its work and was no longer shown in the agents panel |
|
v2.1.234/tui |
Documentation arrivedAnthropic's documentation has since written up /tui, on Claude Code changelog.
the line in the docs* Fixed `/tui` refusing to restart because of an agent-team teammate that had already finished its work and was no longer shown in the agents panel |
|
v2.1.248deniedMcpServers |
Documentation arrivedAnthropic's documentation has since written up deniedMcpServers, on Claude Code changelog.
the line in the docs* Fixed `allowManagedMcpServersOnly`, `deniedMcpServers` and `disableClaudeAiConnectors` set via MDM or `managed-settings.json` being ignored when server-managed settings are also present |
|
v2.1.242deniedMcpServers |
Documentation arrivedAnthropic's documentation has since written up deniedMcpServers, on Claude Code changelog.
the line in the docs* Fixed `allowManagedMcpServersOnly`, `deniedMcpServers` and `disableClaudeAiConnectors` set via MDM or `managed-settings.json` being ignored when server-managed settings are also present |
|
v2.1.271blockReadsOutsideWorkingDirectories |
Documentation arrivedAnthropic's documentation has since written up blockReadsOutsideWorkingDirectories, on Claude Code changelog.
the line in the docs* Fixed Bash commands the permission checker cannot fully analyze skipping the prompt under `permissions.blockReadsOutsideWorkingDirectories`, and a subshell hiding a dangerous `rm` in bypass mode |
|
v2.1.236error_max_budget_usd |
Documentation arrivedAnthropic's documentation has since written up error_max_budget_usd, on Subagents in the SDK.
the line in the docs| Spend | `maxBudgetUsd` in TypeScript, `max_budget_usd` in Python | No limit. Compared against `total_cost_usd`, so subagent requests count | Enforces the cap in three ways: refuses to spawn more subagents, returning `Budget limit reached… |
|
v2.1.268ultracode |
Documentation arrivedAnthropic's documentation has since written up ultracode, on Subagents in the SDK.
the line in the docs| Concurrency | [`CLAUDE_CODE_MAX_CONCURRENT_SUBAGENTS`](/docs/en/env-vars) | `20` subagents running at once, counting every subagent Claude spawns with the Agent tool | Refuses to spawn another subagent, returning `Concurrent subagent lim… |
|
v2.1.271action |
Documentation arrivedAnthropic's documentation has since written up action, on Agent SDK reference - Python.
the line in the docs| `can_use_tool` | [`CanUseTool`](#canusetool) ` \| None` | `None` | Tool permission callback, invoked only when the [permission flow](/docs/en/agent-sdk/permissions#how-permissions-are-evaluated) falls through to a prompt. Not invoked for… |
|
v2.1.271options |
Documentation arrivedAnthropic's documentation has since written up options, on Agent SDK reference - Python.
the line in the docs| `betas` | `list[SdkBeta]` | `[]` | Beta features to enable. See [`SdkBeta`](#sdkbeta) for available options | |
|
v2.1.248system_prompt |
Documentation arrivedAnthropic's documentation has since written up system_prompt, on Agent SDK reference - Python.
the line in the docs| `system_prompt` | `str \| SystemPromptPreset \| SystemPromptFile \| None` | `None` | System prompt configuration. Pass a string for a custom prompt, `{"type": "preset", "preset": "claude_code"}` for Claude Code's system prompt with optio… |
|
v2.1.268can_use_tool |
Documentation arrivedAnthropic's documentation has since written up can_use_tool, on Agent SDK reference - Python.
the line in the docs| `allowed_tools` | `list[str]` | `[]` | Tools to auto-approve without prompting. This does not restrict Claude to only these tools. If you name one of the [task-tracking tools](/docs/en/agent-sdk/todo-tracking#model-availability) here, Cl… |
|
v2.1.251set_model |
Documentation arrivedAnthropic's documentation has since written up set_model, on Agent SDK reference - Python.
the line in the docs| `set_model(model)` | Change the model for the current session. Pass `None` to reset to [Claude Code's default model](/docs/en/model-config) | |
|
v2.1.248set_model |
Documentation arrivedAnthropic's documentation has since written up set_model, on Agent SDK reference - Python.
the line in the docs| `set_model(model)` | Change the model for the current session. Pass `None` to reset to [Claude Code's default model](/docs/en/model-config) | |
|
v2.1.257.claude.json |
Documentation arrivedA small documentation edit on Use Claude Code features in the SDK touched a line naming .claude.json after this was published.
the line in the docsWhen you omit `settingSources`, `query()` reads the same filesystem settings as the Claude Code CLI: user, project, and local settings, CLAUDE.md files, and `.claude/` skills, agents, and commands. To run without these, pass `settingSource… |
|
v2.1.247~/.claude.json |
Documentation arrivedA small documentation edit on Use Claude Code features in the SDK touched a line naming ~/.claude.json after this was published.
the line in the docsWhen you omit `settingSources`, `query()` reads the same filesystem settings as the Claude Code CLI: user, project, and local settings, CLAUDE.md files, and `.claude/` skills, agents, and commands. To run without these, pass `settingSource… |
|
v2.1.238~/.claude.json |
Documentation arrivedA small documentation edit on Use Claude Code features in the SDK touched a line naming ~/.claude.json after this was published.
the line in the docsWhen you omit `settingSources`, `query()` reads the same filesystem settings as the Claude Code CLI: user, project, and local settings, CLAUDE.md files, and `.claude/` skills, agents, and commands. To run without these, pass `settingSource… |
|
v2.1.234~/.claude.json |
Documentation arrivedA small documentation edit on Use Claude Code features in the SDK touched a line naming ~/.claude.json after this was published.
the line in the docsWhen you omit `settingSources`, `query()` reads the same filesystem settings as the Claude Code CLI: user, project, and local settings, CLAUDE.md files, and `.claude/` skills, agents, and commands. To run without these, pass `settingSource… |
|
v2.1.265/clear |
Documentation arrivedA small documentation edit on How the agent loop works touched a line naming /clear after this was published.
the line in the docsWith [streaming input](/docs/en/agent-sdk/streaming-vs-single-mode), a message that is still queued when a turn ends at the max-turns limit stays queued. Claude Code doesn't add it to that turn's last model call. It starts a new turn for t… |
|
v2.1.260/clear |
Documentation arrivedA small documentation edit on How the agent loop works touched a line naming /clear after this was published.
the line in the docsWith [streaming input](/docs/en/agent-sdk/streaming-vs-single-mode), a message that is still queued when a turn ends at the max-turns limit stays queued. Claude Code doesn't add it to that turn's last model call. It starts a new turn for t… |
|
v2.1.257/clear |
Documentation arrivedA small documentation edit on How the agent loop works touched a line naming /clear after this was published.
the line in the docsWith [streaming input](/docs/en/agent-sdk/streaming-vs-single-mode), a message that is still queued when a turn ends at the max-turns limit stays queued. Claude Code doesn't add it to that turn's last model call. It starts a new turn for t… |
|
v2.1.246/clear |
Documentation arrivedA small documentation edit on How the agent loop works touched a line naming /clear after this was published.
the line in the docsWith [streaming input](/docs/en/agent-sdk/streaming-vs-single-mode), a message that is still queued when a turn ends at the max-turns limit stays queued. Claude Code doesn't add it to that turn's last model call. It starts a new turn for t… |
|
v2.1.236error_max_budget_usd |
Documentation arrivedA small documentation edit on How the agent loop works touched a line naming error_max_budget_usd after this was published.
the line in the docsWith [streaming input](/docs/en/agent-sdk/streaming-vs-single-mode), a message that is still queued when a turn ends at the max-turns limit stays queued. Claude Code doesn't add it to that turn's last model call. It starts a new turn for t… |
|
v2.1.271types |
Documentation arrivedAnthropic's documentation has since written up types, on Persist sessions to external storage.
the line in the docsBoth SDK repositories include runnable reference adapters under [`examples/session-stores/`](https://github.com/anthropics/claude-agent-sdk-typescript/tree/main/examples/session-stores) in TypeScript and [`examples/session_stores/`](https:… |
|
v2.1.271types |
Documentation arrivedAnthropic's documentation has since written up types, on Persist sessions to external storage.
the line in the docsBoth SDK repositories include runnable reference adapters under [`examples/session-stores/`](https://github.com/anthropics/claude-agent-sdk-typescript/tree/main/examples/session-stores) in TypeScript and [`examples/session_stores/`](https:… |
|
v2.1.271types |
Documentation arrivedAnthropic's documentation has since written up types, on Persist sessions to external storage.
the line in the docsBoth SDK repositories include runnable reference adapters under [`examples/session-stores/`](https://github.com/anthropics/claude-agent-sdk-typescript/tree/main/examples/session-stores) in TypeScript and [`examples/session_stores/`](https:… |
|
v2.1.248favicon |
Documentation arrivedA small documentation edit on Overview touched a line naming favicon after this was published.
the line in the docsBecause inference runs in your cloud tenant, Claude Desktop on 3P operates inside whatever compliance boundary your provider and region give you. The desktop application itself contacts Anthropic-operated hosts only to download the VM work… |
|
v2.1.248favicon |
Documentation arrivedA small documentation edit on Overview touched a line naming favicon after this was published.
the line in the docsBecause inference runs in your cloud tenant, Claude Desktop on 3P operates inside whatever compliance boundary your provider and region give you. The desktop application itself contacts Anthropic-operated hosts only to download the VM work… |
|
v2.1.267Box |
Documentation arrivedAnthropic's documentation has since written up Box, on Configuration reference.
the line in the docs| `oauth.clientSecret` | `string` | — | Only for IdPs whose token endpoint requires a client secret (e.g. Box). Leave blank for PKCE-only public clients. | |
|
v2.1.238readOnly |
Documentation arrivedAnthropic's documentation has since written up readOnly, on Configuration reference.
the line in the docs| `readOnly` | `boolean` | — | Offer only read tools — the server registers no write tools at all. | |
|
v2.1.239effortLevel |
Documentation arrivedAnthropic's documentation has since written up effortLevel, on Configuration reference.
the line in the docsThe effort level the default model (the first `inferenceModels` entry, or the first model your endpoint returns under discovery) starts at in Chat, Cowork and Code, in place of Anthropic's recommended level for that model: one of `low`, `m… |
|
v2.1.239effortLevel |
Documentation arrivedAnthropic's documentation has since written up effortLevel, on Configuration reference.
the line in the docsThe effort level the default model (the first `inferenceModels` entry, or the first model your endpoint returns under discovery) starts at in Chat, Cowork and Code, in place of Anthropic's recommended level for that model: one of `low`, `m… |
|
v2.1.269/context |
Documentation arrivedA small documentation edit on Or, for the Project Glasswing model with the same capabilities: touched a line naming /context after this was published.
the line in the docs5. **Trim context on the server, or compact in a shape that carries no stale thinking:** If your code truncates or summarizes older turns on the client, the simplest fix is to move that work to server-side [compaction](https://platform.cla… |
|
v2.1.269tee |
Documentation arrivedA small documentation edit on browser-use-tool touched a line naming tee after this was published.
the line in the docsAnthropic has trained the model to resist these prompt injections and has added an extra layer of defense. If you use the browser use tool, classifiers will automatically scan what the browser returns, such as page text or screenshots, to … |
|
v2.1.269tee |
Documentation arrivedA small documentation edit on browser-use-tool touched a line naming tee after this was published.
the line in the docsAnthropic has trained the model to resist these prompt injections and has added an extra layer of defense. If you use the browser use tool, classifiers will automatically scan what the browser returns, such as page text or screenshots, to … |
|
v2.1.269tee |
Documentation arrivedA small documentation edit on browser-use-tool touched a line naming tee after this was published.
the line in the docsAnthropic has trained the model to resist these prompt injections and has added an extra layer of defense. If you use the browser use tool, classifiers will automatically scan what the browser returns, such as page text or screenshots, to … |
|
v2.1.269tee |
Documentation arrivedA small documentation edit on browser-use-tool touched a line naming tee after this was published.
the line in the docsAnthropic has trained the model to resist these prompt injections and has added an extra layer of defense. If you use the browser use tool, classifiers will automatically scan what the browser returns, such as page text or screenshots, to … |
|
v2.1.260disableDeploymentModeChooser |
Documentation arrivedA small documentation edit on Connect Claude Desktop to Claude for Government touched a line naming disableDeploymentModeChooser after this was published.
the line in the docsOnce `bootstrapUrl`, `disableDeploymentModeChooser`, or any other recognized key except the automatic update settings is present in the profile, the device is managed. The in-app configuration window becomes read-only, and locally authored… |
|
v2.1.269Plugins |
Documentation arrivedA small documentation edit on Manage plugins and connectors touched a line naming Plugins after this was published.
the line in the docsA plugin you upload on the **Plugins** card delivers its skills, slash commands, sub-agents, and hooks to members, and its hooks run on the member's machine. A plugin can also declare [MCP servers](/docs/connectors/overview) of its own, an… |
|
v2.1.269Plugins |
Documentation arrivedA small documentation edit on Manage plugins and connectors touched a line naming Plugins after this was published.
the line in the docsA plugin you upload on the **Plugins** card delivers its skills, slash commands, sub-agents, and hooks to members, and its hooks run on the member's machine. A plugin can also declare [MCP servers](/docs/connectors/overview) of its own, an… |