📦 Claude Code v2.1.88 - Computer Use, Fullscreen rendering, Enterprise GitHub

content/CHANGELOG.md

@@ -1,5 +1,49 @@
 # Changelog
 
+## 2.1.88
+
+- Added `CLAUDE_CODE_NO_FLICKER=1` environment variable to opt into flicker-free alt-screen rendering with virtualized scrollback
+- Added `PermissionDenied` hook that fires after auto mode classifier denials — return `{retry: true}` to tell the model it can retry
+- Added named subagents to `@` mention typeahead suggestions
+- Fixed prompt cache misses in long sessions caused by tool schema bytes changing mid-session
+- Fixed nested CLAUDE.md files being re-injected dozens of times in long sessions that read many files
+- Fixed Edit/Write tools doubling CRLF on Windows and stripping Markdown hard line breaks (two trailing spaces)
+- Fixed `StructuredOutput` schema cache bug causing ~50% failure rate in workflows with multiple schemas
+- Fixed memory leak where large JSON inputs were retained as LRU cache keys in long-running sessions
+- Fixed a potential out-of-memory crash when the Edit tool was used on very large files (>1 GiB)
+- Fixed a crash when removing a message from very large session files (over 50MB)
+- Fixed `--resume` crash when transcript contains a tool result from an older CLI version or interrupted write
+- Fixed misleading "Rate limit reached" message when the API returned an entitlement error — now shows the actual error with actionable hints
+- Fixed LSP server zombie state after crash — server now restarts on next request instead of failing until session restart
+- Fixed hooks `if` condition filtering not matching compound commands (`ls && git push`) or commands with env-var prefixes (`FOO=bar git push`)
+- Fixed prompt history entries containing CJK or emoji being silently dropped when they fall on a 4KB boundary in `~/.claude/history.jsonl`
+- Fixed `/stats` losing historical data beyond 30 days when the stats cache format changes
+- Fixed `/stats` undercounting tokens by excluding subagent/fork usage
+- Fixed scrollback disappearing when scrolling up in long sessions
+- Fixed collapsed search/read group badges duplicating in terminal scrollback during heavy parallel tool use
+- Fixed notification `invalidates` not clearing the currently-displayed notification immediately
+- Fixed prompt briefly disappearing after submit when background messages arrived during processing
+- Fixed long `/btw` responses being clipped with no way to scroll — responses now render in a scrollable viewport
+- Fixed Devanagari and other combining-mark text being truncated in assistant output
+- Fixed rendering artifacts on main-screen terminals after layout shifts
+- Fixed voice mode failing to request microphone permission on macOS Apple Silicon
+- Fixed voice push-to-talk not activating for some modifier-combo bindings
+- Fixed voice mode on Windows failing with "WebSocket upgrade rejected with HTTP 101"
+- Fixed Shift+Enter submitting instead of inserting a newline on Windows Terminal Preview 1.25
+- Fixed periodic UI jitter during streaming in iTerm2 when running inside tmux
+- Fixed PowerShell tool incorrectly reporting failures when commands like `git push` wrote progress to stderr on Windows PowerShell 5.1
+- Fixed SDK error result messages (`error_during_execution`, `error_max_turns`) to correctly set `is_error: true` with descriptive messages
+- Fixed task notifications being lost when backgrounding a session with Ctrl+B
+- Fixed PreToolUse/PostToolUse hooks not providing `file_path` as an absolute path for Write/Edit/Read tools
+- Improved PowerShell tool prompt with version-appropriate syntax guidance (5.1 vs 7+)
+- Thinking summaries are no longer generated by default in interactive sessions — set `showThinkingSummaries: true` in settings to restore
+- Auto mode denied commands now show a notification and appear in `/permissions` → Recent tab
+- `/env` now applies to PowerShell tool commands (previously only affected Bash)
+- `/usage` now hides redundant "Current week (Sonnet only)" bar for Pro and Enterprise plans
+- Collapsed tool summary now shows "Listed N directories" for ls/tree/du instead of "Read N files"
+- Image paste no longer inserts a trailing space
+- Pasting `!command` into an empty prompt now enters bash mode, matching typed `!` behavior
+
 ## 2.1.87
 
 - Fixed messages in Cowork Dispatch not getting delivered

content/blog/engineering/claude-code-best-practices.md

@@ -58,6 +58,7 @@ Claude Code overview
 *   Claude Code on the web 
 *   Claude Code on desktop 
 *   [Chrome extension (beta)](https://www.anthropic.com/docs/en/chrome)
+*   [Computer use (preview)](https://www.anthropic.com/docs/en/computer-use)
 *   [Visual Studio Code](https://www.anthropic.com/docs/en/vs-code)
 *   [JetBrains IDEs](https://www.anthropic.com/docs/en/jetbrains)
 *   Code review & CI/CD 

content/blog/engineering/claude-code-sandboxing.md

@@ -68,7 +68,7 @@ Our new sandboxed bash tool and Claude Code on the web offer substantial improve
 To get started with these tools:
 
 1.   Run `/sandbox` in Claude and check out [our docs](https://docs.claude.com/en/docs/claude-code/sandboxing) on how to configure this sandbox.
-2.   Go to [claude.com/code](http://claude.ai/redirect/website.v1.fdeb5e93-0f04-4e8b-a162-49c3e80b18c4/code) to try out Claude Code on the web.
+2.   Go to [claude.com/code](http://claude.ai/redirect/website.v1.00086a34-31b8-49b6-afd7-84f7cb165327/code) to try out Claude Code on the web.
 
 Or, if you're building your own agents, check out our [open-sourced sandboxing code](https://github.com/anthropic-experimental/sandbox-runtime), and consider integrating it into your work. We look forward to seeing what you build.
 

content/claude-code-manifest.json

@@ -1,12 +1,12 @@
 {
   "name": "@anthropic-ai/claude-code",
-  "version": "2.1.87",
+  "version": "2.1.88",
   "author": {
     "name": "Anthropic",
     "email": "support@anthropic.com"
   },
   "license": "SEE LICENSE IN README.md",
-  "_id": "@anthropic-ai/claude-code@2.1.87",
+  "_id": "@anthropic-ai/claude-code@2.1.88",
   "maintainers": [
     {
       "name": "zak-anthropic",
@@ -73,23 +73,22 @@
     "claude": "cli.js"
   },
   "dist": {
-    "shasum": "7727a2e69353a24eb9adf89b6134fd207c086a2a",
-    "tarball": "https://registry.npmjs.org/@anthropic-ai/claude-code/-/claude-code-2.1.87.tgz",
-    "fileCount": 19,
-    "integrity": "sha512-R40p85lv270MsOmuP0VrgZlsBnq6HIoW/aIrPwny6AfhPUmsChPNLkxv/F8Mf6g9iYpGNEfICt5CDw++tKZD0g==",
+    "shasum": "c22a001bea2241defb15d0124939836170389daf",
+    "tarball": "https://registry.npmjs.org/@anthropic-ai/claude-code/-/claude-code-2.1.88.tgz",
+    "fileCount": 20,
+    "integrity": "sha512-ukMtYZCi0I7cD3rt89rnXy20D/Zvk0Gj/SW60xYfz17zTslLz+VuhXw/KPb+2ndp3/ATadJdyNEqdKQlcNk7nQ==",
     "signatures": [
       {
-        "sig": "MEUCIFUWZDQOTg3MYfKwczjITo77nHxO+kAHVab08wmlIuaoAiEA3zSRCOGanzjrSsoq0dD1/EGiUo9kX3WLiIhR0F6tKrU=",
+        "sig": "MEUCIHXjhm3IsYM1o1worSMkPW8rIHqSNsV6D08wJoIArNumAiEAtiif6ZPqt/ovlEXSM0sEc8NuCteOb+yIkDZcff2kFSk=",
         "keyid": "SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"
       }
     ],
-    "unpackedSize": 46604198
+    "unpackedSize": 102754401
   },
   "type": "module",
   "engines": {
     "node": ">=18.0.0"
   },
-  "gitHead": "baec21064a7d4725dd2c34ecab379b04e922973a",
   "scripts": {
     "prepare": "node -e \"if (!process.env.AUTHORIZED) { console.error('ERROR: Direct publishing is not allowed.\\nPlease see the release workflow documentation to publish this package.'); process.exit(1); }\""
   },
@@ -116,7 +115,7 @@
     "@img/sharp-linuxmusl-arm64": "^0.34.2"
   },
   "_npmOperationalInternal": {
-    "tmp": "tmp/claude-code_2.1.87_1774748452104_0.7953736765342561",
+    "tmp": "tmp/claude-code_2.1.88_1774910207554_0.1190643563209497",
     "host": "s3://npm-registry-packages-npm-production"
   }
 }

content/en/about-claude/models/overview.md

@@ -43,6 +43,10 @@ _<sup>2 - **Reliable knowledge cutoff** indicates the date through which a model
 You can query model capabilities and token limits programmatically with the [Models API](/docs/en/api/models/list). The response includes `max_input_tokens`, `max_tokens`, and a `capabilities` object for every available model.
 </Tip>
 
+<Note>
+The Max output values above apply to the synchronous Messages API. On the [Message Batches API](/docs/en/build-with-claude/batch-processing#extended-output-beta), Opus 4.6 and Sonnet 4.6 support up to 300k output tokens by using the `output-300k-2026-03-24` beta header.
+</Note>
+
 <section title="Legacy models">
 
 The following models are still available. Consider migrating to current models for improved performance:
@@ -57,20 +61,18 @@ The following models are still available. Consider migrating to current models f
 | **[Extended thinking](/docs/en/build-with-claude/extended-thinking)** | Yes | Yes | Yes | Yes | Yes | No |
 | **[Priority Tier](/docs/en/api/service-tiers)** | Yes | Yes | Yes | Yes | Yes | No |
 | **Comparative latency** | Fast | Moderate | Moderate | Fast | Moderate | Fast |
-| **Context window** | <Tooltip tooltipContent="~750k words \ ~3.4M unicode characters">1M</Tooltip> (or <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k</Tooltip><sup>1</sup>) tokens | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~750k words \ ~3.4M unicode characters">1M</Tooltip> (or <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k</Tooltip><sup>1</sup>) tokens | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> |
+| **Context window** | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> | <Tooltip tooltipContent="~150k words \ ~680k unicode characters">200k tokens</Tooltip> |
 | **Max output** | 64k tokens | 64k tokens | 32k tokens | 64k tokens | 32k tokens | 4k tokens |
-| **Reliable knowledge cutoff** | Jan 2025<sup>2</sup> | May 2025<sup>2</sup> | Jan 2025<sup>2</sup> | Jan 2025<sup>2</sup> | Jan 2025<sup>2</sup> | <sup>3</sup> |
+| **Reliable knowledge cutoff** | Jan 2025<sup>1</sup> | May 2025<sup>1</sup> | Jan 2025<sup>1</sup> | Jan 2025<sup>1</sup> | Jan 2025<sup>1</sup> | <sup>2</sup> |
 | **Training data cutoff** | Jul 2025 | Aug 2025 | Mar 2025 | Mar 2025 | Mar 2025 | Aug 2023 |
 
 <Warning>
 Claude Haiku 3 (`claude-3-haiku-20240307`) is deprecated and will be retired on April 19, 2026. Migrate to [Claude Haiku 4.5](/docs/en/about-claude/models/overview#latest-models-comparison) before the retirement date. See [model deprecations](/docs/en/about-claude/model-deprecations) for details.
 </Warning>
 
-_<sup>1 - Claude Opus 4.5, Opus 4.1, Opus 4, Haiku 4.5, and earlier models have a 200k-token context window. Claude Sonnet 4.5 and Sonnet 4 default to 200k but can access a [1M-token context window](/docs/en/build-with-claude/context-windows) by including the `context-1m-2025-08-07` beta header. [Long context pricing](/docs/en/about-claude/pricing#long-context-pricing) applies to requests exceeding 200k tokens.</sup>_
-
-_<sup>2 - **Reliable knowledge cutoff** indicates the date through which a model's knowledge is most extensive and reliable. **Training data cutoff** is the broader date range of training data used.</sup>_
+_<sup>1 - **Reliable knowledge cutoff** indicates the date through which a model's knowledge is most extensive and reliable. **Training data cutoff** is the broader date range of training data used.</sup>_
 
-_<sup>3 - Some Haiku models have a single training data cutoff date.</sup>_
+_<sup>2 - Some Haiku models have a single training data cutoff date.</sup>_
 
 </section>
 

content/en/about-claude/models/whats-new-claude-4-6.md

@@ -80,9 +80,11 @@ response = client.beta.messages.create(
 
 [Fine-grained tool streaming](/docs/en/agents-and-tools/tool-use/fine-grained-tool-streaming) is now generally available on all models and platforms. No beta header is required.
 
-### 128k output tokens
+### Higher output token limits
 
-Opus 4.6 supports up to 128k output tokens, doubling the previous 64k limit. This enables longer thinking budgets and more comprehensive responses. The SDKs require streaming for requests with large `max_tokens` values to avoid HTTP timeouts. If you don't need to process events incrementally, use `.stream()` with `.get_final_message()` to get the complete response. See [Streaming Messages](/docs/en/build-with-claude/streaming#get-the-final-message-without-handling-events) for details.
+Opus 4.6 supports up to 128k output tokens. This enables longer thinking budgets and more comprehensive responses. The SDKs require streaming for requests with large `max_tokens` values to avoid HTTP timeouts. If you don't need to process events incrementally, use `.stream()` with `.get_final_message()` to get the complete response. See [Streaming Messages](/docs/en/build-with-claude/streaming#get-the-final-message-without-handling-events) for details.
+
+On the Message Batches API, Opus 4.6 and Sonnet 4.6 can generate up to 300k output tokens by using the `output-300k-2026-03-24` beta header. See [Batch processing](/docs/en/build-with-claude/batch-processing#extended-output-beta) for details.
 
 ### Data residency controls
 

content/en/about-claude/pricing.md

@@ -75,7 +75,7 @@ Prompt caching uses the following pricing multipliers relative to base input tok
 
 Cache write tokens are charged when content is first stored. Cache read tokens are charged when a subsequent request retrieves the cached content. A cache hit costs 10% of the standard input price, which means caching pays off after just one cache read for the 5-minute duration (1.25x write), or after two cache reads for the 1-hour duration (2x write).
 
-These multipliers stack with other pricing modifiers, including the Batch API discount, long context pricing, and data residency.
+These multipliers stack with other pricing modifiers, including the Batch API discount and data residency.
 
 For implementation details, supported models, and code examples, see the [prompt caching documentation](/docs/en/build-with-claude/prompt-caching).
 
@@ -128,44 +128,6 @@ For more information about batch processing, see the [batch processing documenta
 
 Claude Opus 4.6 and Sonnet 4.6 include the full [1M token context window](/docs/en/build-with-claude/context-windows) at standard pricing. (A 900k-token request is billed at the same per-token rate as a 9k-token request.) Prompt caching and batch processing discounts apply at standard rates across the full context window.
 
-For Claude Sonnet 4.5 and Sonnet 4, the 1M token context window is in beta for organizations in [usage tier](/docs/en/api/rate-limits) 4 and organizations with custom rate limits. When the `context-1m-2025-08-07` beta header is included, requests that exceed 200k input tokens are automatically charged at premium long context rates:
-
-| Model | ≤ 200k input tokens<br />Input | ≤ 200k input tokens<br />Output | > 200k input tokens<br />Input | > 200k input tokens<br />Output |
-| --- | --- | --- | --- | --- |
-| Claude Sonnet 4.5 / 4 | \$3 / MTok | \$15 / MTok | \$6 / MTok | \$22.50 / MTok |
-
-Long context pricing for Sonnet 4.5 and Sonnet 4 stacks with other pricing modifiers:
-- The [Batch API 50% discount](#batch-processing) applies to long context pricing
-- [Prompt caching multipliers](#model-pricing) apply on top of long context pricing
-
-<Note>
-Even with the beta flag enabled, requests with fewer than 200k input tokens are charged at standard rates. If your request exceeds 200k input tokens, all tokens incur premium pricing.
-
-The 200k threshold is based solely on input tokens (including cache reads/writes). Output token count does not affect pricing tier selection, though output tokens are charged at the higher rate when the input threshold is exceeded.
-</Note>
-
-For Claude Sonnet 4.5 and Sonnet 4, to check if your API request was charged at premium long context rates, examine the `usage` object in the API response:
-
-```json
-{
-  "usage": {
-    "input_tokens": 250000,
-    "cache_creation_input_tokens": 0,
-    "cache_read_input_tokens": 0,
-    "output_tokens": 500
-  }
-}
-```
-
-Calculate the total input tokens by summing:
-- `input_tokens`
-- `cache_creation_input_tokens` (if using prompt caching)
-- `cache_read_input_tokens` (if using prompt caching)
-
-If the total exceeds 200,000 tokens, the entire request was billed at premium long context rates.
-
-For more information about the `usage` object, see the [API response documentation](/docs/en/api/messages#response-usage).
-
 ### Tool use pricing
 
 Tool use requests are priced based on:

content/en/agent-sdk/python.md

@@ -1190,11 +1190,11 @@ Literal type for SDK beta features.
 SdkBeta = Literal["context-1m-2025-08-07"]
 ```
 
-Use with the `betas` field in `ClaudeAgentOptions` to enable beta features. Use `context-1m-2025-08-07` with Claude Sonnet 4.5 and Sonnet 4 to enable the 1M-token context window.
+Use with the `betas` field in `ClaudeAgentOptions` to enable beta features.
 
-<Note>
-Claude Opus 4.6 and Sonnet 4.6 have a 1M token context window. Including `context-1m-2025-08-07` has no effect on those models.
-</Note>
+<Warning>
+The `context-1m-2025-08-07` beta is retired as of April 30, 2026. Requests using this header with Claude Sonnet 4.5 or Sonnet 4 will return a 400 error. To use a 1M-token context window, migrate to [Claude Sonnet 4.6 or Claude Opus 4.6](/docs/en/about-claude/models/overview), which include 1M context at standard pricing with no beta header required.
+</Warning>
 
 ### `McpSdkServerConfig`
 

content/en/agent-sdk/typescript.md

@@ -282,7 +282,7 @@ Configuration object for the `query()` function.
 | `agents` | `Record<string, [`AgentDefinition`](#agent-definition)>` | `undefined` | Programmatically define subagents |
 | `allowDangerouslySkipPermissions` | `boolean` | `false` | Enable bypassing permissions. Required when using `permissionMode: 'bypassPermissions'` |
 | `allowedTools` | `string[]` | `[]` | Tools to auto-approve without prompting. This does not restrict Claude to only these tools; unlisted tools fall through to `permissionMode` and `canUseTool`. Use `disallowedTools` to block tools. See [Permissions](/docs/en/agent-sdk/permissions#allow-and-deny-rules) |
-| `betas` | [`SdkBeta`](#sdk-beta)`[]` | `[]` | Enable beta features (e.g., `['context-1m-2025-08-07']`) |
+| `betas` | [`SdkBeta`](#sdk-beta)`[]` | `[]` | Enable beta features |
 | `canUseTool` | [`CanUseTool`](#can-use-tool) | `undefined` | Custom permission function for tool usage |
 | `continue` | `boolean` | `false` | Continue the most recent conversation |
 | `cwd` | `string` | `process.cwd()` | Current working directory |
@@ -2118,13 +2118,9 @@ Available beta features that can be enabled via the `betas` option. See [Beta he
 type SdkBeta = "context-1m-2025-08-07";
 ```
 
-| Value | Description | Compatible Models |
-|:------|:------------|:------------------|
-| `'context-1m-2025-08-07'` | Enables the 1 million token [context window](/docs/en/build-with-claude/context-windows). | Claude Sonnet 4.5, Claude Sonnet 4 |
-
-<Note>
-Claude Opus 4.6 and Sonnet 4.6 have a 1M token context window. Including `context-1m-2025-08-07` has no effect on those models.
-</Note>
+<Warning>
+The `context-1m-2025-08-07` beta is retired as of April 30, 2026. Requests using this value with Claude Sonnet 4.5 or Sonnet 4 will return a 400 error. To use a 1M-token context window, migrate to [Claude Sonnet 4.6 or Claude Opus 4.6](/docs/en/about-claude/models/overview), which include 1M context at standard pricing with no beta header required.
+</Warning>
 
 ### `SlashCommand`