anthropics/claude-code
Coding-Agent InfrastructureClaude Code
v2.1.231
v2.1.231- What's changed
- Fixed MCP OAuth sign-in failing with a redirect URI mismatch for servers that use a pre-registered OAuth client, such as Slack
anthropics/claude-code
Open repository ↗Claude Code
952 signals · 959 observation events
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
PROJECT NEWS
All news →Release, roadmap, and discussion
Observation trail
- changedmetrics
- changedupdatedAt, labels
- changedupdatedAt, labels
- changedupdatedAt
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- changedupdatedAt, labels
- changedupdatedAt, labels
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- discoveredinitial snapshot
- changedmetrics
- changedmetrics
- discoveredinitial snapshot
- changedupdatedAt, metrics
- discoveredinitial snapshot
- discoveredinitial snapshot
- changedupdatedAt, metrics
- discoveredinitial snapshot
- changedupdatedAt, metrics
- discoveredinitial snapshot
- changedupdatedAt
- discoveredinitial snapshot
issue
[FEATURE] /simplify should cut AI slop comments
### Preflight Checklist - [x] I have searched [existing requests](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20label%3Aenhancement) and this feature hasn't been requested yet - [x] This is a single feature request (not multiple features) ### Problem Statement AI loves to overcomment the code. ### Proposed Solution The /simplify command nee
issue
Scheduled tasks re-prompt for tool permissions on automatic/cron fires despite prior manual approval
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? Summary: Schedule
issue
[BUG] Web version leaks hook systemMessage into user chat UI (should be Claude-only context)
## Description In the Claude Code **web version** (claude.ai/code), `systemMessage` output from `UserPromptSubmit` hooks is displayed in the user-facing chat UI. According to the hooks documentation, exit-0 stdout — including `systemMessage` — should be injected as context for Claude only, not shown to the user. The **terminal CLI** behaves correctly (silent
issue
Feedback submission (/bug) crashes long-running sessions — payload appears to scale with session size, killing background subagents
## Summary In a long-running session, submitting feedback via the in-app feedback flow (`/bug`) reproducibly crashes the entire Claude Code process. It happened three times today in the same session, each attempt within seconds of submitting. The apparent mechanism is memory: the box shows a large allocation spike on submit, and the process dies without an e
issue
[BUG][Cowork] Escape on the Settings window auto-denies the pending tool permission dialog behind it
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? ## Environment -
issue
[BUG] Claude Desktop MSIX installation fails with HRESULT 0x80073CFF on Windows 11 Pro despite all preconditions passing — matches #60319
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? Claude Desktop in
issue
[BUG] Windows: Grep corrupts forward slashes to backslashes in context lines when paths render relative
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? On Windows, the `
issue
[Feature Request] Expose remaining budget and a configurable override for the 1000-agent-per-workflow lifetime cap
The Workflow tool caps total `agent()` calls at 1000 per run as a runaway-loop backstop, documented as "set far above any real workflow." For a corpus-scale workflow (e.g. per-partition fetch/judgment passes fanned out across many ecosystems/families, each itself dispatching further per-item sub-calls), that assumption doesn't always hold — a source corpus c
issue
[BUG] CLaude desktop just completely uninstalled itself from my windows PC?? And all my work is lost!
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? First of all how
issue
[Bug] Claude ignores project context and memories, requires constant redirection
**Bug Description** Claude entirely overlooked memories and otehr conventions along with project details to embark onto designing it all from scratch. this latest iteration of the model is akin to a 5 years old kid that needs constant pro-active attention as it constantly digress and diversify itsefl, not staying on task... i am starting to get really tired
issue
Tool-captured screenshots are re-sent every turn and dominate transcript bytes
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? Screenshots captu
issue
pool
**Bug Description** 내 서버 만드는데 멍청하게 굶 **Environment Info** - Platform: win32 - Terminal: pycharm - Version: 2.1.231 - Feedback ID: 15fe1512-9521-4f97-98c8-37cb22ee150d **Errors** ```json [] ```
issue
[Bug] Repeated false positive safety flags when creating topic lists with approved CVP
**Bug Description** this is stupid. Second time in 10 min I'm flagged for trying to create a list of topics for a soc. We have an approved CVP although that's worth basically nothing. **Environment Info** - Platform: darwin - Terminal: iTerm.app - Version: 2.1.231 - Feedback ID: 0abb87bb-365b-4ff7-a062-5cd1e44a9d8c **Errors** ```json [] ```
issue
[BUG] Desktop auto-update fails with 0x80073D02 while app is running, leaves MSIX package Modified/NeedsRemediation — app unlaunchable (twice in one week)
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? Twice in one week
issue
code-review skill: one nested background finder agent's task-notification silently dropped; parent has no fallback and returns fabricated status instead of its required output
## Summary Running the built-in `code-review` skill at `high` effort against a PR, the skill's own review agent fanned out into 4 background finder agents (via the `Agent` tool, each spawned async/in-background). 3 of the 4 completed and delivered their `<task-notification>` normally. The 4th agent also completed successfully (confirmed from its own transcri
issue
[Bug] API Error: Empty or Malformed Response with HTTP 200 Status
**Bug Description** API Error: API returned an empty or malformed response (HTTP 200) — check for a proxy or gateway intercepting the request **Environment Info** - Platform: darwin - Terminal: Apple_Terminal - Version: 2.1.231 - Feedback ID: 2733c243-32d4-4fc9-9a67-17d47f3b89a0 **Errors** ```json [{"error":"Error: API returned an empty or malformed response
issue
Desktop app: "Failed to update marketplace" for plugins installed via remote/account-scoped install path
## Description Refreshing a plugin in the Desktop app that was installed through the remote/account-scoped install path fails with "Failed to update marketplace". The Desktop app shells out to the CLI to perform the update, but the CLI has no local record of the marketplace, so the command fails. ## Steps to reproduce 1. Install a plugin via the Desktop app
issue
HTTP MCP: valid token treated as expired — "OAuth session expired and could not be refreshed" when provider issues no refresh token (mcp.facebook.com/ads)
## Summary A remote HTTP MCP server (Meta's official ads server at `https://mcp.facebook.com/ads`, registered with `--client-id`) stopped connecting. The surface error is a 30s timeout, but `--debug` reveals the real one: ``` Failed to authenticate: OAuth session expired and could not be refreshed ``` The stored access token is **valid and not expired** (iss
issue
[BUG] Multiple Claude Desktop actions (deleting active chat, opening Preferences) freeze the entire macOS system
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? ### Summary Certa
issue
[FEATURE] Import of old chats (history)
### Preflight Checklist - [x] I have searched [existing requests](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20label%3Aenhancement) and this feature hasn't been requested yet - [x] This is a single feature request (not multiple features) ### Problem Statement Due to a enum bug I could not work anymore with the app only with CLI, the aold c
issue
Windows desktop 1.28929.0 (MSIX): whole app exits on GPU process crash, triggered by session preview/Browser pane
### What happens On Windows 11 the desktop app (MSIX build 1.28929.0) dies several times a day — window disappears, no error dialog, all open sessions lost. There is no OS-level crash record: no `Application Error` (event ID 1000) for `Claude.exe` in the Windows Application log, and no Crashpad dump is produced. In `%APPDATA%\Claude\logs\main.log` every sing
issue
[BUG] `/goal`: the Stop-hook condition is frozen at creation time and the verifier cannot see later cancellation
### Preflight Checklist - [x] I have searched existing issues and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code — the sessions analysed here ran on 2.1.224/2.1.226 and 2.1.228 > **Disclosure.** This report was drafted by Claude Code (Opus 5)
issue
Prohibited block forbids all deletion, unscoped — blocks rm, log scrubbing, and cleanup in coding/ops agents
**Bug Description** The `Prohibited` safety block forbids "permanently deleting data" outright, with no override even on explicit instruction from the operator. In a coding/ops agent this removes `rm`, log scrubbing, and any cleanup of the agent's own artifacts — work that is routine, non-negotiable, and frequently *security* work. The block reads as a compu
issue
[Bug] Unexpected language switching to Portuguese regardless of configuration
**Bug Description** Claude Code switches reply language mid-session despite CLAUDE.md forcing français, no user trigger. **Environment Info** - Platform: linux - Terminal: pycharm - Version: 2.1.231 - Feedback ID: 7d7b577b-2511-48c9-9a44-bbe07e186b2c Hypothesis: repo's git history has a contributor whose name is from another language (surfaced via git log/bl
issue
[MODEL] Template not replicating
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Amodel) for similar behavior reports - [x] This report does NOT contain sensitive information (API keys, passwords, etc.) ### Type of Behavior Issue Claude ignored my instructions or configuration ### What You
issue
[BUG] `/goal clear` typed while the agent is working is treated as chat text and never executed
### Preflight Checklist - [x] I have searched existing issues and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code — the sessions analysed here ran on 2.1.226 and 2.1.228 > **Disclosure.** This report was drafted by Claude Code (Opus 5) analysi
issue
[BUG] `/goal` and `/goal clear` report "No goal set" after a met/failed verdict while the Stop hook stays armed
### Preflight Checklist - [x] I have searched existing issues and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code — the sessions analysed here ran on 2.1.224/2.1.226 and 2.1.228 > **Disclosure.** This report was drafted by Claude Code (Opus 5)
issue
Agent should stop and report, not work around, a failed Docker context connection
**Summary** While working on a Dockerized project, I asked Claude Code to inspect a Docker image (to check a file-ownership UID) using the currently active Docker context, which talks to a remote engine over SSH. That connection timed out (an intermittent/expected condition in this environment). Instead of stopping and reporting the failed connection, the ag
issue
Claude generates AL/Business Central code that misuses [TryFunction] assuming it rolls back transactions
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Amodel) for similar behavior reports - [x] This report does NOT contain sensitive information (API keys, passwords, etc.) ### Type of Behavior Issue Other unexpected behavior ### What You Asked Claude to Do Sum
issue
[Bug] Classifiers not working during XDR evaluation despite document submission
**Bug Description** even im enrolled in cyber policies i alredy submitted my documents still again and again im facing this issue XDR evalution im doing right now still the classifiers not working **Environment Info** - Platform: win32 - Terminal: windows-terminal - Version: 2.1.228 - Feedback ID: e9ed10a9-fce2-4427-a9c9-5b7c8cb7ede2 **Errors** ```json [{"er
issue
[BUG] Desktop app (macOS): scheduled sessions with background tasks leak harness processes indefinitely (task-notification race, follow-up to #39632)
## Summary On the macOS desktop app, every scheduled/automated session that uses background work (background Bash, Monitor, subagents) leaves its harness process alive indefinitely after its last turn completes. The processes accumulate daily. The transcripts of every affected session end with `queue-operation: enqueue` entries carrying `<task-notification>`
issue
Desktop app hangs previewing localhost dev server; uninstall/reinstall silently wipes all session transcripts
**Environment** - OS: Windows 11 Home 10.0.26200 - Claude desktop app: 1.28929.0 (MSIX / Microsoft Store install) **Bug 1: App hangs when the embedded Browser pane opens a localhost dev server tab** While using the Browser pane tools (preview_start/navigate) to check a Next.js dev server on `localhost:3000`, the entire desktop app stopped responding (not jus
issue
Remote control of a Claude Code session requires being physically at the host machine to enable
**Problem:** Remote control of a Claude Code session (via the iPhone app / claude.ai) currently needs a slash command run locally, on the host machine itself, for each session before it can be remote-controlled. If you didn't run that command on a given session before leaving the machine, there's no way to enable it remotely later — you have to be physically
issue
[BUG] Print mode (-p --output-format stream-json): session ends with result: success while background tasks are running and a ScheduleWakeup is pending — scheduled wakeup can never fire, background tasks are killed
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? ## Environment -
issue
MCP channel notifications with list- or int-valued params.meta fields are silently dropped (no error, connection stays healthy)
## Environment - Claude Code 2.1.221 (Windows, native), MCP server over stdio (Python), channels enabled via `--dangerously-load-development-channels` - Still apparently unfixed: CHANGELOG entries 2.1.222–2.1.229 contain no channel/meta/notification fix (not re-tested on latest; happy to re-verify on request) ## Summary `notifications/claude/channel` notific
issue
[FEATURE] Desktop app: keyboard shortcut for archiving a session
## Description The Code tab has no keyboard shortcut for archiving a session. Archive is reachable only by hovering the session row in the sidebar and clicking the archive icon, or by asking Claude to do it (which raises an approval card in every permission mode, by design). The documented shortcut table (`Cmd+/`, and https://code.claude.com/docs/en/desktop#
issue
[BUG] VS Code Extension on OSX - Home/End keys don't move to end of line
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? In the chat windo
issue
Auto mode classifier denies tool calls in bypassPermissions sessions (v2.1.231)
### Environment - **Claude Code**: 2.1.231 - **OS**: macOS (Darwin 25.5.0) - **Launched with**: `claude --dangerously-skip-permissions` - **Settings**: `~/.claude/settings.json` → `permissions.defaultMode: "bypassPermissions"` (set explicitly, so the CLI flag and the settings file agree — no ambiguity about the intended mode) - **Status bar**: `bypass permis
issue
[MODEL] Title: Model ignores explicit global scope-of-file-access rule when searching for a file — repeats the violation even while building a fix for it
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Amodel) for similar behavior reports - [x] This report does NOT contain sensitive information (API keys, passwords, etc.) ### Type of Behavior Issue Claude accessed files outside the working directory ### What
issue
[Bug] Tool destroyed files during font color swap operation due to unsafe clone-to-save optimization
**Bug Description** tool destroyedf a whole collecttion of 555 files wuth swapping over fonts in wrong colors, i spend says doind and redoing beucase this piece of shit tool is set to not read clone rto savw cpu, legal case registration only, eil bynch of dickgwads **Environment Info** - Platform: darwin - Terminal: Apple_Terminal - Version: 2.1.226 - Feedba
issue
[Bug] Anthropic API Error: Safety filter blocking legitimate security scanning tasks
**Bug Description** I was scanning my own code for vulneratiblities and got this message: "Fable 5's safeguards flagged this message. Our intentionally broad safeguards allow us to deliver more capabilities faster, but can sometimes flag legitimate coding, cybersecurity, and biology tasks. Switched to Opus 4.8. Send feedback with /feedback or learn more" **E
issue
Queued cross-session message lands cleanly but doesn't trigger a new agent turn until something else flushes the queue
## Summary A message delivered via cross-session messaging (`send_message` / the CCD session-management MCP tools) can land cleanly in the target session's transcript as a real `cross-session-message` entry, but the receiving session does not automatically start a new turn to process it. The message just sits there — session status stays `running`/idle — unt
issue
[Bug] Tool cannot color font and adapt space during clone operation
**Bug Description** tool cant co,or font and adaptspace to clone. you call this shit exectuion... **Environment Info** - Platform: darwin - Terminal: Apple_Terminal - Version: 2.1.226 - Feedback ID: 1f4a6f34-7964-4c24-9bff-52d1c5bc6dbe **Errors** ```json [] ```
issue
[Bug] Fable 5 declines explicitly assigned work and insists on delegating it to another agent, immediately after weekly-allowance waste from recovery loops
## Summary During a long-running Claude Code session (VS Code native extension, Windows 11, Claude Fable 5), I observed a third recurring failure signature, related to but distinct from the ones I reported in #84757 (unsolicited closing posture) and #82126 (interlocutor-specific correctness collapse): **The model declines to perform an explicitly assigned im
issue
[BUG] MCP tool array/list parameters get silently stringified mid-call, intermittently — breaks any tool expecting List[str]
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? ## Summary When c
issue
[Feature Request] Option to use git branch name as the session title (desktop app has no way to automate this)
### Summary Add a way — ideally a setting — to have a session's title/name in the session list default to the **current git branch name**, particularly in the **desktop app** where there is no launch-time equivalent of `claude -n`. ### Problem When working in one repo across many sessions, the auto-generated titles converge on near-identical summaries (e.g.
issue
[BUG] Claude Code repeatedly ignores my documented (and re-stated) rules about not triggering needless permission prompts
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? **Environment:**
issue
[BUG] MCP progress notifications are received but discarded once a call is auto-backgrounded at 120s
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? When an MCP tool
issue
[sonnet and opus] Inability to use web search in dependency research
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Amodel) for similar behavior reports - [x] This report does NOT contain sensitive information (API keys, passwords, etc.) ### Type of Behavior Issue Claude accessed files outside the working directory ### What
issue
[BUG] Can not bypass GIT on Windows Desktop
### Preflight Checklist - [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) and this hasn't been reported yet - [x] This is a single bug report (please file separate reports for different bugs) - [x] I am using the latest version of Claude Code ### What's Wrong? The windows deskt