Source Intelligence

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

All releases Home olderv2.1.133 v2.1.137newer

Claude Code v2.1.136

12 entries read diff v2.1.133 → v2.1.136 Markdown

Claude Code 2.1.136 adds admin-controlled managed settings helpers, a new autoMode.hard_deny rule category, and gradual-rollout routine notifications. It also improves MCP configuration compatibility, --worktree failure messages, WSL image paste support, and PowerShell safety checks.

Find
Pick an entry · j / k steps through
8 entries

Changesopen

Enterprise Managed Settings Helper#

What

Organizations can now configure an admin-controlled helper executable that computes managed settings at startup.

Usage
{
  "policyHelper": {
    "path": "/usr/local/bin/claude-policy-helper",
    "timeoutMs": 5000,
    "refreshIntervalMs": 60000
  }
}
Details
  • The helper is honored only from admin-controlled policy sources.
  • It can return managedSettings, claudeMd, and appendSystemPrompt.
  • Claude Code validates the helper path, rejects non-absolute paths, and can refresh the helper output periodically.
  • If refresh fails, Claude Code retains the current policy instead of dropping managed settings.
Evidence

Managed settings helper schema and runtime execution (search for "policyHelper" and "Executable that computes managed settings at startup. Honored only from admin-controlled policy sources.")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Auto Mode Hard-Deny Rules#

What

Auto mode custom rules now distinguish soft blocks from hard security blocks.

Usage
{
  "autoMode": {
    "allow": ["$defaults"],
    "soft_deny": ["$defaults"],
    "hard_deny": [
      "$defaults",
      "Never send secrets to external endpoints"
    ],
    "environment": ["$defaults"]
  }
}
Details
  • soft_deny is now described as destructive or irreversible behavior that clear user intent can authorize.
  • hard_deny is new and covers security-boundary actions that user intent does not clear.
  • claude auto-mode defaults, claude auto-mode config, and claude auto-mode critique now understand all four categories: allow, soft_deny, hard_deny, and environment.
Evidence

Auto mode schema and CLI output now include hard_deny (search for "Rules for the auto mode classifier HARD BLOCK section" and "autoMode.{allow, soft_deny, hard_deny, environment}")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Related

Other releases about the same thing. Found by shared names or similar wording; neither means one caused the other.

Routine Fired Notifications [Gradual Rollout]#

What

Claude Code can now notify you when remote routines have run.

Usage
claude
Details
  • When enabled, Claude Code checks Claude routine trigger state and shows a notification such as a routine name or “routines ran”.
  • Notifications link back to /code/routines.
  • This is gated by remote-session support and a feature flag, so not all users will see it immediately.
  • Organizations can disable routine dispatch through policy.
Evidence

Routine trigger notification path (search for "routine-fired", "/code/routines", and "Routines are disabled by your organization's policy.")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

MCP Server Config Compatibility#

Claude Code now accepts streamable-http as an MCP transport alias for HTTP and gives clearer errors for invalid MCP server types.

Usage
{
  "mcpServers": {
    "example": {
      "type": "streamable-http",
      "url": "https://example.com/mcp"
    }
  }
}
Evidence

MCP transport normalization and validation (search for "streamable-http" and "Valid types are: stdio, sse, http (or streamable-http), ws, sdk")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Related

Other releases about the same thing. Found by shared names or similar wording; neither means one caused the other.

Safer MCP Client Secret Storage#

When adding HTTP/SSE MCP servers with --client-secret, Claude Code now preserves the server config even if secure storage is unavailable, then tells you how to retry secret storage.

Usage
claude mcp add --transport http my-server https://example.com/mcp --client-id abc --client-secret
Evidence

MCP add fallback warning (search for "Server added, but the client secret could not be stored")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Clearer Worktree Setup Failures#

--worktree now gives more actionable errors when workspace trust has not been accepted, when a worktree branch is already checked out elsewhere, or when a stale worktree directory blocks reuse.

Usage
claude --worktree feature-name
Evidence

Worktree diagnostics (search for "Workspace trust not yet accepted. Run \claude\ once in this directory and accept the trust dialog, then retry with --worktree." and "is already checked out in a worktree at")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

WSL Clipboard Image Paste#

Claude Code can now fall back to Windows PowerShell from WSL when checking or saving image clipboard contents.

Evidence

WSL clipboard fallback to Windows PowerShell (search for "powershell.exe)" and "Get-Clipboard -Format Image")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Related

Other releases about the same thing. Found by shared names or similar wording; neither means one caused the other.

Broader /effort max Documentation#

The /effort help text now documents max as available for Opus 4.6/4.7 and Sonnet 4.6, instead of only Opus 4.6/4.7.

Usage
/effort max
Evidence

Effort help text (search for "Opus 4.6/4.7, Sonnet 4.6")

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

1 entry

Bug Fixesopen

#

  • Fixed --worktree starting before workspace trust is accepted, replacing a later failure with a direct trust message. Evidence: trust gate added before worktree creation (search for "Workspace trust not yet accepted")
  • Improved PowerShell 5.1 command safety detection for cwd-first command shadowing. Evidence: new warning for commands that create a file which shadows a later command (search for "Windows PowerShell 5.1 cwd-first resolution")
  • Improved PowerShell validation around runtime-resolved Unicode codepoint escapes. Evidence: new classifier message (search for "PowerShell \u{HEX} codepoint escape is runtime-resolved and cannot be statically validated."`)
  • Fixed bare Git repository detection to recognize .git files that contain gitdir: and avoid misclassifying linked worktrees or nested repositories. Evidence: updated Git structure detection (search for "gitdir:")
Verbatim
Official · Anthropic

Anthropic’s official release notes

Published verbatim by Anthropic for v2.1.136. Text is unmodified from the upstream changelog. Everything else on this page came out of the bundle instead, which is why the two lists don't match.

  • Added CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTEL to re-enable the session quality survey for enterprises capturing responses through OpenTelemetry
  • Added settings.autoMode.hard_deny for auto mode classifier rules that block unconditionally regardless of user intent or allow exceptions
  • Fixed MCP servers configured in .mcp.json, plugins, and claude.ai connectors silently disappearing after /clear in the VS Code extension, JetBrains plugin, and Agent SDK
  • Fixed a rare login loop where a concurrent credential write could overwrite a freshly-rotated OAuth token and force re-login
  • Fixed MCP OAuth refresh tokens being lost when multiple servers refresh concurrently — users with several remote MCP servers should no longer need daily re-authentication
  • Fixed an API error (400) when extended thinking emitted a redacted thinking block after a tool call
  • Fixed --resume / --continue not finding sessions when the project path contains underscores
  • Fixed plan mode not blocking file writes when a matching Edit(...) allow rule exists
  • WSL2: image paste from Windows clipboard now works via a PowerShell fallback when xclip/wl-paste cannot read image data
  • Fixed plugin Stop/UserPromptSubmit hooks failing when cache cleanup deletes a version still in use by a running session
  • Improved visual consistency across slash command dialogs: standardized footer hints, dialog spacing, and arrow-key styling, and the dialog frame now appears immediately during loading instead of popping in after
  • Fixed colors appearing at wrong positions in bash command output and markdown code blocks
  • Fixed ReasonML diffs rendering corrupted "undefined" text artifacts at word-diff boundaries
  • Fixed worktree exit dialog warning about uncommitted files in the wrong directory after worktree removal
  • Fixed @ file picker not matching files created mid-session in small non-git directories
  • Fixed @-mention file picker not finding files in directories with more than 100 entries
  • Fixed failed tool calls not being click-to-expand in fullscreen mode when their output was truncated
  • Fixed Backspace and Ctrl+Backspace getting swapped after using Ctrl+G to open an external editor on terminals with persistent extended-key modes
  • Fixed /usage weekly reset showing time of day instead of the calendar date
  • Fixed welcome banner ellipsis causing column overflow on CJK terminals
  • Fixed /insights crash when session history contains tool calls with malformed input fields
  • Fixed a renderer crash when a tool's collapsibility classification changes mid-session
  • Fixed a skills entry in plugin.json hiding the plugin's default skills/ directory, and listing a file path now shows an error instead of failing silently
  • Fixed IDE shell-integration lock files not respecting CLAUDE_CONFIG_DIR
  • Fixed trailing whitespace in copied terminal output during streaming
  • Fixed plugin uninstall and enable/disable not matching slugs case-insensitively
  • Fixed tool error truncation marker showing a negative count for surrogate-pair strings
  • Fixed env vars from CLAUDE_ENV_FILE SessionStart hooks going stale after /resume or /clear
  • Fixed /branch saving a multi-line session title when given a pasted multi-line name
  • Fixed a stray leading space on the second line of wrapped text at the column boundary
  • Fixed Esc not dismissing dialogs in /install-github-app, /desktop, /resume, and /web-setup
  • Fixed /doctor MCP schema errors not naming the missing field or showing the source file path
  • Fixed Bash permission prompts showing an internal parser diagnostic instead of a user-readable explanation
  • Fixed plugin slash commands with spaces (e.g. /myplugin review) not resolving to their namespaced form
  • Fixed AskUserQuestion discarding multi-select answers when supplied as an array
  • Fixed /clear <name> not labeling the cleared session for /resume
  • Fixed CronList output missing qualifiers and the scheduled prompt
  • Fixed "Jump to bottom" overlay leaving color artifacts on CJK characters in fullscreen mode
  • Fixed wide markdown tables leaving a stale bordered render in terminal scrollback while streaming
  • Fixed pasted text being silently dropped when a long prompt with a pasted-text placeholder was auto-truncated
  • Fixed /release-notes getting stuck on an old version after a failed changelog refresh
  • Fixed /mcp server list not scrolling when there are more servers than fit in the terminal
  • Fixed mid-input slash command autocomplete not working after an initial slash command
  • Fixed scrolling to bottom re-engaging auto-follow with autoScrollEnabled: false
  • Fixed prompt suggestions being auto-submitted by Enter on an empty input instead of requiring Tab or arrow to accept
  • Fixed keyboard shortcut hints not reflecting rebound keys from keybindings.json
  • Fixed /settings language change being reverted on Escape after confirming
  • Fixed /terminal-setup only appearing in autocomplete on exact name match instead of partial prefixes
  • Fixed "Chat about this" on an AskUserQuestion dialog erasing the question text
  • Fixed MCP tool results being invisible when the server returns content blocks
  • Improved error message when --worktree collides with an existing or stale worktree
  • Changed plugin marketplace removal key to d (matching delete elsewhere) instead of r which collided with retry
System prompt

No change to the system prompt since v2.1.133.

Claude Code, interactive mode