Claude Code v2.1.214
Released Jul 18, 2026 (1 month ago).
132 issues reference this build
87 still open
45 resolved (34%)
"References this build" means the issue body mentions running 2.1.214 (from the
bug-report template). It reflects report volume, not a proven defect count — an issue may mention a
version without being caused by it. First-party data from the anthropics/claude-code tracker.
Top issues reported on v2.1.214
Ranked by community engagement (reactions and comments).
- MCP servers declaring draft-07 outputSchema are entirely unusable: "unsupported dialect" rejected client-side before dispatch
- Fable 5 safeguard flag silently reroutes sessions to Opus 4.8 for hundreds of turns with no persistent indicator
- Task tools AND TodoWrite both hidden in interactive sessions by live tengu_vellum_ash experiment (newest models); env override ineffective
- [FEATURE] Sessions should follow a renamed/moved project folder (persistent project identity)
- [BUG] Cannot update payment method — "connection to link account has been closed"; support (Fin) reports account as Free while plan is Max 5x
- [Bug] Fable model inaccessible despite included in max plan and available usage credits
- [FEATURE] Routines have no durable in-app notification surface — findings only survive by being written into Calendar/Gmail
- [BUG] VS Code extension: "Could not locate the Claude CLI on PATH" false positive since v2.1.214 (regression from v2.1.212), likely related to where.exe output handling on Windows with non-ASCII username
- [FEATURE] Desktop app: 'folder missing' section + bulk re-link for sessions whose project folder no longer exists
- [Bug] Escape key denies tool execution instead of exiting btw tool
- Bundled workflows (code-review, deep-research) offer no model override — all 10-15 subagents inherit the session model
- Session-limit interactive prompt blocks queued scheduled prompts (CronCreate//loop) in unattended sessions
- Background Bash task completion never resumes in-process subagents — they idle forever while finished work sits on disk
- Subagent `maxTurns` frontmatter is not honored (agent runs past the declared cap)
- [Bug] Claude Code performance regression: Task execution significantly slower than previous runs
- TCC network-volume permission re-prompts on every update (binary path includes version number)
- Bundled ugrep allocates 4-17 GB to search a 64 KB file when the pattern has a trailing `.{N}` bound (Linux/WSL2, 2.1.214)
- [BUG] Daemon bg/spare sessions leave title-only .jsonl stubs in the resume picker; resuming errors, then starts a blank session that re-submits the last prompt
- [FEATURE] Expose agentType per task in the subagentStatusLine payload
- [BUG] MCP dict/object tool arguments serialized as JSON strings (regression of #3084) — 422 on object-typed params in 2.1.214
- I don't have enough technical details to generate a proper GitHub issue title. The report "its okay idk why fable is tripping" lacks: - What "fable" refers to (a feature, module, or dependency) - What error or behavior is occurring - When/how it happens -
- [Bug] Agent removal fails when worktree directory deleted out-of-band
- Inline --agents 'tools: []' no longer denies tools in headless -p mode (2.1.214); no flag combination behaviorally denies all
- Claude Code blocked by "weekly limit reached" (resets Jul 25) while all usage meters show 0% — claude.ai web works on same account
- [Bug] Anthropic API Safeguard False Positive on Academic Content in Markdown Editorial Review
- MCP server connects successfully but its tools are never exposed to the model
- Bash permissions.deny rules not enforced in subagent (Agent/Task tool) shells — denied command executes, only post-hoc security warning
- CLAUDE_CODE_OAUTH_TOKEN from claude setup-token returns 401 Invalid bearer token on every API call
- [Bug] Mixed language characters in Claude response (Chinese character in English text)
- Background daemon restart drops in-flight agent workers without respawn (bg adopt: respawned=0 dead=N)
Changelog
What's changed
- Fixed single-segment
dir/**allow rules likeEdit(src/**)auto-approving writes to nesteddir/directories anywhere in the tree instead of only<cwd>/dir - Fixed a permission-check bypass affecting commands run in Windows PowerShell 5.1 sessions
- Fixed Bash permission checks to fail closed on file-descriptor redirect forms that bash parses differently than the permission analyzer
- Fixed Bash permission checks misjudging very long commands — commands over 10,000 characters now always prompt instead of running automatically
- Fixed Bash permission checks treating zsh variable subscripts and modifiers in
[[ ]]comparisons as inert text — these commands now prompt for approval - Fixed Bash permission checks to no longer auto-approve certain
helpandmancommands that could run unsafe options, command substitutions, or backslash paths - Fixed permission prompts on remote sessions that could proceed before the local confirmation dialog
- Added the EndConversation tool: Claude can end sessions with highly abusive users or jailbreak attempts, as on claude.ai since 2025 — see https://www.anthropic.com/research/end-subset-conversations
- Added a periodic progress heartbeat for long-running tool calls that previously went silent
- Added an ISO
modifiedtimestamp to memory file frontmatter - Added
message.uuid,client_request_id, andtool_sourceattributes to OpenTelemetry log events for message-level correlation and tool provenance - Added
CLAUDE_CODE_OTEL_CONTENT_MAX_LENGTHto configure the 60 KB truncation limit on OpenTelemetry content attributes - Added reasoning effort to the
subagentStatusLinepayload, so custom agent rows can render model and effort - Added permission prompts for
dockercommands (including the Podmandockershim) carrying daemon-redirect flags (--url,--connection,--identity, and Podman's remote mode) that previously ran without one - Fixed a crash when a GrowthBook feature evaluates to null, and a bug where a malformed flag payload could wipe the cached feature flags
- Fixed Bash tool killing the Claude session when a
pkill -fpattern accidentally matched the CLI's own process (Linux) - Fixed unbounded memory growth when
--settingspoints at a device file or multi-GB file; oversized (>2 MiB) settings files now fail at startup with a clear error - Fixed streaming turns failing with "Socket is closed" behind corporate proxies on Windows
- Fixed stream-json output truncation at exit for slow-reading SDK/pipeline consumers; the exit drain now scales with queued bytes instead of a flat 2s cap
- Fixed scheduled tasks refusing their own configured prompt as untrusted input — the fired prompt is now delivered as the session's assigned task
- Fixed PowerShell tool commands hanging until timeout when a child process waited on standard input (Windows)
- Fixed Python scripts under the PowerShell tool crashing with UnicodeDecodeError when reading non-UTF-8 data from standard input (Windows)
- Fixed Python scripts run via the PowerShell tool crashing with UnicodeEncodeError on non-ASCII output, and PowerShell 7 error messages containing raw ANSI escape sequences (Windows)
- Fixed the PowerShell tool reporting
where.exe,fc.exe, anddiff.exeas errors when they return a valid negative answer (Windows) - Fixed
>and>>under the PowerShell tool on Windows PowerShell 5.1 writing UTF-16LE files that other tools couldn't read as UTF-8 - Fixed a displaced background daemon deleting its successor's control socket on shutdown, which made the next client kill the healthy replacement daemon
- Fixed background sessions parked with
←or/backgroundand left idle keeping the background daemon and a worker process alive indefinitely - Fixed completed background sessions being impossible to remove via
claude rmor the agent view once the background service had gone idle - Fixed background sessions dispatched from a non-git folder being impossible to delete from t