90 pages for experimental.

beta-headers Claude Developer Platform · in the page api/beta-headers …ps://platform.claude.com/docs/en/api/beta-headers description: Access experimental features before they become part of the standard API with the `anthropic-beta` header or the SDK's `betas` parameter.… routines-fire Claude Developer Platform · in the page api/claude-code/routines-fire …by sending an authenticated POST request. --- <Warning> This is an experimental API. Request and response shapes, rate limits, and token semantics might change. </Warning> [Claude Code](https://co… apple-foundation-models Claude Developer Platform · in the page cli-sdks-libraries/libraries/apple-foundation-models …thand that guesses): ```swift let model = ClaudeModel( id: "claude-experimental-x", capabilities: .init(samplingParams: false, effortLevels: [.low, .high]) ) ClaudeLanguageModel(name: model, auth… overview Claude Developer Platform · in the page release-notes/overview …/8606378-how-do-i-use-the-workbench) in the Claude Help Center. * The experimental prompt tools APIs for generating, improving, and templatizing prompts (`/v1/experimental/generate_prompt`, `/v1/exper… Escalate hard decisions with the advisor tool Claude Code CLI · in the page advisor …ults at key moments during a task. > [!NOTE] > > The advisor tool is experimental and requires the Anthropic API. It is not available on Amazon Bedrock, Claude Platform on AWS, Google Cloud's Agent P… Modifying system prompts Claude Code CLI · in the page agent-sdk/modifying-system-prompts …g one string. The same happens whenever you set [`CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1`](/docs/en/llm-gateway-protocol#disable-pre-release-capabilities). To split the prompt, import `SYSTEM_PROMP… Work with sessions Claude Code CLI · in the page agent-sdk/sessions …ess") { console.log(message.result); } } ``` > [!NOTE] > > The experimental [V2 session API](/docs/en/agent-sdk/typescript-v2-preview.md), which provided `createSession()` with a `send` / `stre… Scale to many tools with tool search Claude Code CLI · in the page agent-sdk/tool-search …e loaded into context on every turn. | Setting [`CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS`](/docs/en/env-vars) keeps tool search off. You can't override it by setting `ENABLE_TOOL_SEARCH` yourself. On… Agent SDK reference - TypeScript Claude Code CLI · in the page agent-sdk/typescript …| number; permissionMode?: PermissionMode; criticalSystemReminder_EXPERIMENTAL?: string; }; ``` | Field | Required | Description | | :- | :- | :- | | `description` | Yes | Natural language descri… TypeScript SDK V2 session API (removed) Claude Code CLI · in the page agent-sdk/typescript-v2-preview …erence if you maintain code on Agent SDK 0.2.x or earlier. V2 was an experimental session API that removed the need for async generators and yield coordination. Instead of managing generator state ac… Orchestrate teams of Claude Code sessions Claude Code CLI · in the page agent-teams …saging, and centralized management. > [!WARNING] > > Agent teams are experimental and disabled by default. Enable them by setting `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` in your [settings.json](/doc… Run agents in parallel Claude Code CLI · in the page agents …with a shared task list and inter-agent messaging, managed by a lead. Experimental and disabled by default | You want Claude to split a project into pieces, assign them, and keep the workers in sync |… Auto mode classifier request charges Claude Code CLI · in the page auto-mode-classifier-billing …requires Claude Code v2.1.281 or later. Setting `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` while `CLAUDE_CODE_AUTO_MODE_SERVER` is unset turns the server's checks off as well, except as [Disable pre-… Best practices for Claude Code Claude Code CLI · in the page best-practices …watch them from one screen * [Agent teams](/docs/en/agent-teams.md): experimental and disabled by default. Automated coordination of multiple sessions with shared tasks, messaging, and a team lead B… Claude Code changelog Claude Code CLI · in the page changelog …reen reader again on every spinner frame * Fixed `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS` not removing the structured-output format from session-title and prompt-hook requests, which Bedrock-backed ga… Channels reference Claude Code CLI · in the page channels-reference …channel — Claude Code registers a listener for it capabilities: { experimental: { 'claude/channel': {} } }, // Claude Code delivers this to Claude as context when the server connects, so it kn… Use Claude Code in the cloud Claude Code CLI · in the page claude-code-on-the-web …eams.md) are off by default but can be enabled by adding `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` to your [environment variables](/docs/en/cloud-environments.md#set-environment-variables). ### Take b… Explore the .claude directory Claude Code CLI · in the page claude-directory …d`, `effort`, `isolation`, `color`, `initialPrompt`, `omitClaudeMd`, `experimental` | [Subagent frontmatter](/docs/en/sub-agents#supported-frontmatter-fields) | | `output-styles/*.md` | `name`, `descr… Manage costs effectively Claude Code CLI · in the page costs …session ends. * Agent teams are disabled by default. Set `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` in your [settings.json](/docs/en/settings.md) or environment to enable them. See [enable agent teams](… Environment variables Claude Code CLI · in the page env-vars …| | `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS` | Set to `1` to strip Anthropic-specific `anthropic-beta` request headers and beta tool-schema… Error reference Claude Code CLI · in the page errors …hat gateways must forward. * As a fallback, set [`CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1`](/docs/en/env-vars) before launching. [Disable pre-release capabilities](/docs/en/llm-gateway-protocol.md#di… Glossary Claude Code CLI · in the page glossary …indow and you can interact with any of them directly. Agent teams are experimental and disabled by default; see [Enable agent teams](/docs/en/agent-teams.md#enable-agent-teams). Learn more: [Run agen… Hooks reference Claude Code CLI · in the page hooks …lob to verify conditions before returning a decision. Agent hooks are experimental and may change. See [Agent-based hooks](#agent-based-hooks). When an event fires, its matching hooks run in parallel… Automate actions with hooks Claude Code CLI · in the page hooks-guide …: "agent"`: multi-turn verification with tool access. Agent hooks are experimental and may change. See [Agent-based hooks](#agent-based-hooks). ### Combine results from multiple hooks When multiple… Connect Claude Code to an LLM gateway Claude Code CLI · in the page llm-gateway-connect …e Code sends to Anthropic-format endpoints | Set `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1`, which suppresses most pre-release fields; see [feature pass-through](/docs/en/llm-gateway-protocol.md#featu… Claude Code gateway compatibility guide Claude Code CLI · in the page llm-gateway-protocol …eature-pass-through), unless the developer sets [`CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS`](#disable-pre-release-capabilities) | The subset Amazon Bedrock and Agent Platform accept | | Request fields f… Roll out an LLM gateway for your organization Claude Code CLI · in the page llm-gateway-rollout …u want developers' pickers populated from it | | `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS` | Stops Claude Code sending pre-release capability headers and body fields. [Disable pre-release capabilities]… Connect Claude Code to tools via MCP Claude Code CLI · in the page mcp …our other projects. Use local scope for personal development servers, experimental configurations, or servers with credentials you don't want in version control. > [!NOTE] > > The term "local scope"… Choose a permission mode Claude Code CLI · in the page permission-modes …that doesn't have it yet, unless you've also set `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1`. If you set `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` and leave `CLAUDE_CODE_AUTO_MODE_SERVER` unset, Claud… Test plugins with evals Claude Code CLI · in the page plugin-evals …on the command line for a single run: * **In `plugin.json`**: add `"experimental": { "evals": "quality/evals" }`. * **On the command line**: pass `--eval-dir quality/evals` to both `claude plugin ev… Plugin commands reference Claude Code CLI · in the page plugins/cli-reference …` | Directory below the plugin that holds the cases | The manifest's `experimental.evals`, else `evals` | | `--json [path]` | Print the [result document](/docs/en/plugin-evals.md#json-result) to stdou… Add components to a plugin Claude Code CLI · in the page plugins/components …nd`, `omitClaudeMd`, `isolation`, `color`, and the `cacheTtl` key of `experimental`. The only valid `isolation` value is `"worktree"`. See [supported frontmatter fields](/docs/en/sub-agents.md#support… Plugin manifest reference Claude Code CLI · in the page plugins/manifest-reference …}, "lspServers": "./.lsp.json", "outputStyles": "./styles/", "experimental": { "themes": "./themes/", "monitors": "./config/monitors.json" }, "userConfig": { "api_token": {… Marketplace reference Claude Code CLI · in the page plugins/marketplace-reference …ill be ignored at load time.` | Warning | `plugins[i].metadata` | | `'experimental' must be an object containing component declarations; got <type>. It will be ignored at load time.` | Warning | `plug… How Claude Code uses prompt caching Claude Code CLI · in the page prompt-caching …on every provider and connection unless you set [`CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS`](/docs/en/llm-gateway-protocol#disable-pre-release-capabilities), in which case that block is sent uncached.… Automate work with scheduled tasks Claude Code CLI · in the page routines …-H "Authorization: Bearer sk-ant-oat01-xxxxx" \ -H "anthropic-beta: experimental-cc-routine-2026-04-01" \ -H "anthropic-version: 2023-06-01" \ -H "Content-Type: application/json" \ -d '{"text"… Configure the sandboxed Bash tool Claude Code CLI · in the page sandboxing …tname and, by default, does not terminate or inspect TLS traffic. The experimental [`network.tlsTerminate`](/docs/en/settings-reference#sandbox-network-tlsterminate) setting makes the built-in proxy t… All settings Claude Code CLI · in the page settings-reference …terminate TLS so it can read the contents of HTTPS requests. This is experimental, and `mask` [credential substitution](/docs/en/sandboxing.md#mask-credentials) requires it. Set `{}` to generate an e… Create custom subagents Claude Code CLI · in the page sub-agents …, `omitClaudeMd`, and `isolation`. * **Ignored fields**: `color` and `experimental` aren't accepted here and are ignored rather than rejected. For what Claude Code does with a value it can't load, an… Week 25 · June 15–19, 2026 Claude Code CLI · in the page whats-new/2026-w25 …nstead of showing "Connection closed while thinking"With `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` set, every session has one implicit team, so you spawn teammates directly with the Agent tool's `name`… Changelog Claude Documentation · in the page cowork/changelog …ments that set no custom base URL to no longer suppress Claude Code's experimental features, so tool search is on by default there (on Vertex AI with Claude 4.5 and newer models) and `toolSearchEnable… Plugin pre-submission checklist Claude Documentation · in the page plugins/pre-submission-checklist …m/docs/en/plugins/manifest-reference) does, and keep them out of the `experimental` object | Blocks | | | Set `description`, `author`, and `version` | Warning | | ### README and license The director… Configuration reference Claude Documentation · in the page third-party/claude-desktop/configuration …e with no custom base URL**: not needed. The app leaves Claude Code's experimental betas on there, as terminal Claude Code does, so tool search is on by default (on Vertex AI, for Claude 4.5 and newer… Deploy Claude Desktop on 3P with an LLM gateway Claude Documentation · in the page third-party/claude-desktop/gateway …ktop turns tool search off by default, along with Claude Code's other experimental beta features, because strict gateways reject the experimental `anthropic-beta` request headers and request fields th… Tool Annotations Charter Model Context Protocol · in the page community/interest-groups/tool-annotations …ng SDK features or reference implementations (may be explored in the [experimental repo](https://github.com/modelcontextprotocol/experimental-ext-tool-annotations) but is not the IG's primary purpose)… SDK Tiering System Model Context Protocol · in the page community/sdk-tiers …orted SDKs with complete protocol implementation, including all non-experimental features and optional capabilities like sampling and elicitation * **Tier 2**: Actively-maintained SDKs working towar… Interceptors Charter Model Context Protocol · in the page community/working-groups/interceptors …60 minutes | Technical discussion, proposal review | ## Resources * Experimental extension repository: [modelcontextprotocol/experimental-ext-interceptors](https://github.com/modelcontextprotocol/ex… Progressive Disclosure Charter Model Context Protocol · in the page community/working-groups/progressive-disclosure …s**: maintaining a shared repository of reference implementations and experimental findings ### Out of Scope * **Approving spec changes**: this WG does not have authority to approve protocol changes… Server Card Charter Model Context Protocol · in the page community/working-groups/server-card …om/channels/1358869848138059966/1399986204405141534) ## Resources * Experimental extension repository: [modelcontextprotocol/experimental-ext-server-card](https://github.com/modelcontextprotocol/exp… Skills Over MCP Charter Model Context Protocol · in the page community/working-groups/skills-over-mcp …p to Working Group | | 2026-04-14 | Initial charter (formalized from [experimental-ext-skills](https://github.com/modelcontextprotocol/experimental-ext-skills) repo README, which served as the de fact…