docs: Claude Code v2.1.158 - plugins auto-load from .claude/skills no marketplace needed + auto mode on Bedrock/Vertex/Foundry

content/CHANGELOG.md

@@ -1,5 +1,45 @@
 # Changelog
 
+## 2.1.158
+
+- Auto mode is now available on Bedrock, Vertex, and Foundry for Opus 4.7 and Opus 4.8. Opt in by setting `CLAUDE_CODE_ENABLE_AUTO_MODE=1`
+
+## 2.1.157
+
+- Plugins in `.claude/skills` directories are now automatically loaded, no marketplace required
+- Added `claude plugin init <name>` to scaffold a new plugin in `.claude/skills`
+- Added autocomplete for `/plugin` arguments: subcommands, installed plugin names, and plugins from known marketplaces
+- `claude agents`: the `agent` field in `settings.json` is now honored for dispatched sessions, with `--agent <name>` to override it
+- `EnterWorktree` can now switch between Claude-managed worktrees mid-session
+- `tool_decision` telemetry events now include `tool_parameters` (bash commands, MCP/skill names) when `OTEL_LOG_TOOL_DETAILS=1`
+- Worktrees managed by Claude are now left unlocked when the agent finishes, so `git worktree remove`/`prune` can clean them up
+- Fixed unprocessable images (zero-byte, corrupt) attached via paste, MCP, or dialog crashing the request instead of becoming a text placeholder
+- Fixed sandbox network permission prompts appearing in auto and bypass-permissions mode when using the desktop app, IDE extensions, or SDK
+- Fixed `claude agents` completed sessions not retiring when an idle subagent was still parked or had leaked a backgrounded shell
+- Fixed `claude agents` pressing Esc not cancelling a slow "opening…", leaving the list unresponsive
+- Fixed background agent worktrees under `.claude/worktrees/` being orphaned after the 30-day job retention sweep
+- Fixed background sessions re-attached after a sleep/wake not telling the model the correct date
+- Fixed copy-on-select in `claude agents` not reaching the system clipboard inside tmux with `set-clipboard on` (regression in 2.1.153)
+- Fixed `--resume` not reporting background subagents that were running when the previous Claude Code process exited
+- Fixed the `--resume` session picker leaving its contents on the terminal after exiting in fullscreen mode
+- Fixed `--worktree` and `--worktree --tmux` returning to the canonical repo root instead of the current linked worktree
+- Fixed the `/model` picker showing an incorrect "Newer version available" hint when the selected model is already the newest in its family; the pinned-model row now shows the model's description instead of its raw ID
+- Fixed literal markdown markers (backticks, asterisks) appearing in the in-progress message text in fullscreen mode
+- Fixed the terminal freezing after approving the managed-settings security dialog at startup
+- Fixed a rare duplicate line appearing in scrollback after the terminal UI redraws
+- Fixed right-click paste duplicating the clipboard in the VS Code, Cursor, and Windsurf integrated terminals
+- WSL: fixed image paste (`alt+v` keybinding), screenshot paste on Windows 11, and added support for dragging images from Windows Explorer
+- Improved performance of long and resumed conversations by eliminating redundant message-rendering recomputations
+- `/terminal-setup` now disables GPU acceleration in VS Code/Cursor/Windsurf integrated terminals to prevent garbled-text rendering
+- The Feature of the Week credit-claim status now appears as a notification in the status area instead of a line above the prompt
+- `claude agents`: slash-command autocomplete in the dispatch input now matches substrings
+- Removed the "bash commands will be sandboxed" startup banner — sandbox status still shows in `/status` and when a command is blocked
+- Removed the "/ide for …" startup hint toast
+- [IDE] Fixed clicking Stop while a background subagent is running not actually stopping it
+- [VSCode] Fixed the fast mode indicator not appearing on Opus 4.8
+- Pressing backspace right after a workflow trigger keyword now dismisses the workflow request (same as alt+w) instead of deleting a character
+- Added a "Workflow keyword trigger" setting in /config to stop the word "workflow" in a prompt from triggering a dynamic workflow
+
 ## 2.1.156
 
 - Fixed an issue when using Opus 4.8 where thinking blocks were modified, leading to API errors.

content/claude-code-manifest.json

@@ -1,12 +1,12 @@
 {
   "name": "@anthropic-ai/claude-code",
-  "version": "2.1.156",
+  "version": "2.1.158",
   "author": {
     "name": "Anthropic",
     "email": "support@anthropic.com"
   },
   "license": "SEE LICENSE IN README.md",
-  "_id": "@anthropic-ai/claude-code@2.1.156",
+  "_id": "@anthropic-ai/claude-code@2.1.158",
   "maintainers": [
     {
       "name": "zak-anthropic",
@@ -73,20 +73,20 @@
     "claude": "bin/claude.exe"
   },
   "dist": {
-    "shasum": "32f21eb881e84f421195873842c68c367093d43e",
-    "tarball": "https://registry.npmjs.org/@anthropic-ai/claude-code/-/claude-code-2.1.156.tgz",
+    "shasum": "210d119f4954d45b91aa0be8a756ec283bd6d500",
+    "tarball": "https://registry.npmjs.org/@anthropic-ai/claude-code/-/claude-code-2.1.158.tgz",
     "fileCount": 7,
-    "integrity": "sha512-DRIqsawy+n+LtNBaxOW+3JYLaehbCdEdc+mZjYv/zRnZ1bHeTetJBcV41TagNjL00hHjrlALdl76wmA4s/PVQQ==",
+    "integrity": "sha512-K6FjD5lQs1TEP4q3SvJalixywOzRgFEHXAgRiRKjJjvFDY41bvpwFDfuZYdi4foAycDunMKVFiy973IOUJXWWg==",
     "signatures": [
       {
-        "sig": "MEUCIAz7iPG3QRZaIOiHKoXT1QSJwRGEkNW2fLQJeEHoeTU8AiEAgtiITFRAZCekOQ17GEVlOht7C4javPiJRal5Ii4AJDI=",
+        "sig": "MEUCIA8f5zLoHfQAzGsnPUdh8BZh6DtGGW+TYcpHoHRTtW/RAiEAuNOVBcrGSKoD/Npz62+9UosWv0sPRi925bAToSipqvI=",
         "keyid": "SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"
       }
     ],
-    "unpackedSize": 145690
+    "unpackedSize": 145798
   },
   "type": "module",
-  "_from": "file:staged-npm/anthropic-ai-claude-code-2.1.156.tgz",
+  "_from": "file:staged-npm/anthropic-ai-claude-code-2.1.158.tgz",
   "engines": {
     "node": ">=18.0.0"
   },
@@ -98,26 +98,27 @@
     "name": "wolffiex",
     "email": "wolffiex@anthropic.com"
   },
-  "_resolved": "/home/runner/work/claude-cli-internal/claude-cli-internal/staged-npm/anthropic-ai-claude-code-2.1.156.tgz",
-  "_integrity": "sha512-DRIqsawy+n+LtNBaxOW+3JYLaehbCdEdc+mZjYv/zRnZ1bHeTetJBcV41TagNjL00hHjrlALdl76wmA4s/PVQQ==",
+  "_resolved": "/home/runner/work/claude-cli-internal/claude-cli-internal/staged-npm/anthropic-ai-claude-code-2.1.158.tgz",
+  "_integrity": "sha512-K6FjD5lQs1TEP4q3SvJalixywOzRgFEHXAgRiRKjJjvFDY41bvpwFDfuZYdi4foAycDunMKVFiy973IOUJXWWg==",
   "_npmVersion": "11.13.0",
   "description": "Use Claude, Anthropic's AI assistant, right from your terminal. Claude can understand your codebase, edit files, run terminal commands, and handle entire workflows for you.",
   "directories": {},
   "_nodeVersion": "24.16.0",
   "dependencies": {},
   "_hasShrinkwrap": false,
+  "readmeFilename": "README.md",
   "optionalDependencies": {
-    "@anthropic-ai/claude-code-linux-x64": "2.1.156",
-    "@anthropic-ai/claude-code-win32-x64": "2.1.156",
-    "@anthropic-ai/claude-code-darwin-x64": "2.1.156",
-    "@anthropic-ai/claude-code-linux-arm64": "2.1.156",
-    "@anthropic-ai/claude-code-win32-arm64": "2.1.156",
-    "@anthropic-ai/claude-code-darwin-arm64": "2.1.156",
-    "@anthropic-ai/claude-code-linux-x64-musl": "2.1.156",
-    "@anthropic-ai/claude-code-linux-arm64-musl": "2.1.156"
+    "@anthropic-ai/claude-code-linux-x64": "2.1.158",
+    "@anthropic-ai/claude-code-win32-x64": "2.1.158",
+    "@anthropic-ai/claude-code-darwin-x64": "2.1.158",
+    "@anthropic-ai/claude-code-linux-arm64": "2.1.158",
+    "@anthropic-ai/claude-code-win32-arm64": "2.1.158",
+    "@anthropic-ai/claude-code-darwin-arm64": "2.1.158",
+    "@anthropic-ai/claude-code-linux-x64-musl": "2.1.158",
+    "@anthropic-ai/claude-code-linux-arm64-musl": "2.1.158"
   },
   "_npmOperationalInternal": {
-    "tmp": "tmp/claude-code_2.1.156_1779999172460_0.1439017309709112",
+    "tmp": "tmp/claude-code_2.1.158_1780103543475_0.7705489452989578",
     "host": "s3://npm-registry-packages-npm-production"
   }
 }

content/en/agents-and-tools/mcp-connector.md

@@ -23,6 +23,14 @@ This feature is **not** eligible for [Zero Data Retention (ZDR)](/docs/en/build-
 - **OAuth authentication**: Support for OAuth Bearer tokens for authenticated servers
 - **Multiple servers**: Connect to multiple MCP servers in a single request
 
+## When Claude uses MCP tools
+
+Once an MCP server is connected, Claude calls its tools when the user's request maps to a tool's described capability, either explicitly ("search Jira for open bugs") or implicitly ("what's blocking the release?" with a Jira server attached).
+
+Claude does **not** call an MCP tool for general knowledge questions about a connected service. Asking "how do Notion databases work?" with a Notion server attached is answered directly; asking "what's in my Projects database?" triggers the tool.
+
+You can steer how readily Claude calls MCP tools through your system prompt. See [When Claude uses tools](/docs/en/agents-and-tools/tool-use/overview#when-claude-uses-tools) for general guidance and example phrasings.
+
 ## Limitations
 
 - Of the feature set of the [MCP specification](https://modelcontextprotocol.io/introduction#explore-mcp), only [tool calls](https://modelcontextprotocol.io/docs/concepts/tools) are currently supported.
@@ -454,7 +462,7 @@ Set `enabled: false` as the default, then explicitly enable specific tools:
 
 ### Denylist: disable specific tools
 
-Enable all tools by default, then explicitly disable unwanted tools:
+Enable all tools by default, then explicitly disable unwanted tools. Denylisting write or destructive tools is recommended when building read-only assistants, or when you want a human confirmation step before state changes:
 
 ```json
 {
@@ -585,6 +593,8 @@ You can connect to multiple MCP servers by including multiple server definitions
 }
 ```
 
+With many tools available, Claude selects based on tool names and descriptions. Clear, specific tool descriptions improve selection accuracy. For large tool sets (dozens of tools across several servers), consider enabling [`defer_loading`](#tool-configuration-options) with the [Tool search tool](/docs/en/agents-and-tools/tool-use/tool-search-tool) so only relevant tools are surfaced per query.
+
 ## Authentication
 
 For MCP servers that require OAuth authentication, you'll need to obtain an access token. The MCP connector beta supports passing an `authorization_token` parameter in the MCP server definition.

content/en/agents-and-tools/remote-mcp-servers.md

@@ -20,6 +20,10 @@ To connect to a remote MCP server:
 
 For more information about using remote MCP servers with the Claude API, see the [MCP connector docs](/docs/en/agents-and-tools/mcp-connector).
 
+<Note>
+Once connected, remote MCP tools follow the same triggering behavior as any other tool. See [When Claude uses MCP tools](/docs/en/agents-and-tools/mcp-connector#when-claude-uses-mcp-tools).
+</Note>
+
 ## Remote MCP server examples
 
 <MCPServersTable platform="mcpConnector" />

content/en/agents-and-tools/tool-use/advisor-tool.md

@@ -582,7 +582,7 @@ If your agent exposes other planner-like tools (for example, a todo list tool),
 
 #### Suggested system prompt for coding tasks
 
-For coding tasks where you want consistent advisor timing and around two to three calls per task, prepend the following blocks to your executor system prompt before any other sentences that mention the advisor. On internal coding evaluations this pattern produced the highest intelligence at near-Sonnet cost.
+Without system-prompt steering, the executor tends to under-call the advisor in some domains — coding tasks in particular. For coding tasks where you want consistent advisor timing and around two to three calls per task, prepend the following blocks to your executor system prompt before any other sentences that mention the advisor.
 
 Timing guidance:
 

content/en/agents-and-tools/tool-use/computer-use-tool.md

@@ -743,6 +743,7 @@ Here are some tips on how to get the best quality outputs:
 4. For repeatable tasks or UI interactions, include example screenshots and tool calls of successful outcomes in your prompt.
 5. If you need the model to log in, provide it with the username and password in your prompt inside XML tags such as `<robot_credentials>`. Using computer use within applications that require login increases the risk of bad outcomes as a result of prompt injection. Review [Mitigate jailbreaks and prompt injections](/docs/en/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks) before providing the model with login credentials.
 6. When constructing a user turn's `content` array, place the instruction text *before* the screenshot image. Providing the target description before the image is processed improves click accuracy.
+7. When using `computer_20251124` with `enable_zoom: true` set, Claude zooms in on a region when asked about small text or specific UI elements that aren't legible at the screenshot's default resolution, such as file names in a sidebar, tab titles, status-bar text, line numbers, or button labels. If Claude isn't zooming when you expect, ask about a specific region or element rather than the screen as a whole.
 
 <Tip>
   If you repeatedly encounter a clear set of issues or know in advance the tasks

content/en/agents-and-tools/tool-use/overview.md

@@ -77,6 +77,18 @@ Tool access is one of the highest-leverage primitives you can give an agent. On
 
 ---
 
+## When Claude uses tools
+
+With the default `tool_choice` of `{"type": "auto"}`, Claude decides on each turn whether to call a tool or respond directly. It calls a tool when the request maps to that tool's described capability and the answer isn't already in context; it responds directly for stable knowledge, creative tasks, and conversational turns.
+
+This boundary is steerable through your system prompt. If Claude isn't calling tools when you expect, a light instruction like `"Use the tools to investigate before responding."` measurably increases tool use; a stronger form like `"Always call a tool first before responding."` pushes further. Conversely, `"Use your judgment about whether to call a tool or respond directly."` keeps triggering behavior conservative.
+
+For a hard guarantee rather than a nudge, use [`tool_choice`](/docs/en/agents-and-tools/tool-use/define-tools#forcing-tool-use).
+
+Each server tool's page describes its own trigger boundary in more detail. See for example [the web search tool](/docs/en/agents-and-tools/tool-use/web-search-tool) or [the code execution tool](/docs/en/agents-and-tools/tool-use/code-execution-tool).
+
+---
+
 ## Tool use examples
 
 For a complete hands-on walkthrough, see the [tutorial](/docs/en/agents-and-tools/tool-use/build-a-tool-using-agent). For reference examples of individual concepts, see [Define tools](/docs/en/agents-and-tools/tool-use/define-tools) and [Handle tool calls](/docs/en/agents-and-tools/tool-use/handle-tool-calls).

content/en/agents-and-tools/tool-use/web-fetch-tool.md

@@ -44,6 +44,15 @@ When you add the web fetch tool to your API request:
 The web fetch tool currently does not support websites dynamically rendered with JavaScript.
 </Note>
 
+### When Claude fetches
+
+Claude fetches when the request points at a specific page or document:
+
+- A URL is provided in the conversation (or a previous tool result)
+- The user names a specific resource (a particular article, README, pricing page, or documentation section) without a URL, and the [web search tool](/docs/en/agents-and-tools/tool-use/web-search-tool) is also enabled so Claude can locate it first (see [Combined search and fetch](#combined-search-and-fetch))
+
+Claude does **not** fetch for general-knowledge or open-ended questions that don't reference a specific page. "Summarize this article: `<url>`" triggers a fetch; "what are best practices for REST API design?" is answered directly.
+
 ### Dynamic filtering
 
 Fetching full web pages and PDFs can quickly consume tokens, especially when only specific information is needed from large documents. With the `web_fetch_20260209` tool version, Claude can write and execute code to filter the fetched content before loading it into context.
@@ -691,6 +700,8 @@ In this workflow, Claude will:
 3. Use web fetch to retrieve full content
 4. Provide detailed analysis with citations
 
+When both the web search and web fetch tools are enabled, and the user names a specific page or document without providing a URL (for example, "read the README from the anthropics/anthropic-sdk-python repository"), Claude uses web search to locate it, then fetches the result.
+
 ## Prompt caching
 
 For caching tool definitions across turns, see [Tool use with prompt caching](/docs/en/agents-and-tools/tool-use/tool-use-with-prompt-caching).

content/en/build-with-claude/adaptive-thinking.md

@@ -309,7 +309,7 @@ You can combine adaptive thinking with the [effort parameter](/docs/en/build-wit
 |:-------------|:------------------|
 | `max` | Claude always thinks with no constraints on thinking depth. Available on Claude Opus 4.8, Claude Mythos Preview, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4.6. |
 | `xhigh` | Claude always thinks deeply with extended exploration. Available on Claude Opus 4.8 and Claude Opus 4.7. |
-| `high` (default) | Claude always thinks. Provides deep reasoning on complex tasks. |
+| `high` (default) | Claude almost always thinks. Provides deep reasoning on complex tasks. |
 | `medium` | Claude uses moderate thinking. May skip thinking for very simple queries. |
 | `low` | Claude minimizes thinking. Skips thinking for simple tasks where speed matters most. |
 
@@ -804,6 +804,16 @@ will meaningfully improve answer quality — typically for problems
 that require multi-step reasoning. When in doubt, respond directly.
 ```
 
+To encourage thinking instead, use a phrase like:
+
+```text
+This task involves multi-step reasoning. Think carefully before responding.
+```
+
+Steering effectiveness can be sensitive to exact wording — if one phrasing doesn't produce the behavior you want, try a more direct variant.
+
+You can also steer thinking on a per-message basis from the user turn. Appending `"Please think hard before responding."` to a user message encourages Claude to think on that turn; `"Answer directly without deliberating."` suppresses it. This works independently of the system prompt and is useful when only some requests in a conversation warrant extended reasoning.
+
 <Warning>
 Steering Claude to think less often may reduce quality on tasks that benefit from reasoning. Measure the impact on your specific workloads before deploying prompt-based tuning to production. Consider testing with lower [effort levels](/docs/en/build-with-claude/effort) first.
 </Warning>

content/en/build-with-claude/context-windows.md

@@ -93,6 +93,8 @@ Claude 4 models support [interleaved thinking](/docs/en/build-with-claude/extend
 For more information about using tools with extended thinking, see the [extended thinking guide](/docs/en/build-with-claude/extended-thinking#extended-thinking-with-tool-use).
 </Note>
 
+Claude's tool selection is designed to hold with large input documents — choosing the right tool (or correctly abstaining) when the conversation includes 100K+ tokens of non-tool context. For reducing context consumed by the tools themselves, see [Manage tool context](/docs/en/agents-and-tools/tool-use/manage-tool-context), or defer tool definitions with the [tool search tool](/docs/en/agents-and-tools/tool-use/tool-search-tool).
+
 <NextOpus />, [Claude Mythos Preview](https://anthropic.com/glasswing), Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4.6 have a 1M-token context window on the Claude API, Amazon Bedrock, and Vertex AI. On Microsoft Foundry, <NextOpus /> has a 200k-token context window. Other Claude models, including Claude Sonnet 4.5 and Sonnet 4 (deprecated), have a 200k-token context window.
 
 A single request can include up to 600 images or PDF pages (100 for models with a 200k-token context window). When sending many images or large documents, you may approach [request size limits](/docs/en/api/overview#request-size-limits) before the token limit.