Claude Code v2.1.214 Released: OTel Tracing, Permission Fixes, EndConversation Tool, and Docker Protection

Anthropic pushed Claude Code v2.1.214 with a heavy emphasis on permission security, telemetry improvements, and Windows PowerShell fixes. Here are the practical takeaways for anyone running the CLI.
Permission Checks Tightened
Bash fixes:
- Single-segment
dir/**allow rules (e.g.Edit(src/**)) no longer auto-approve writes to nesteddir/directories anywhere in the tree — only under<cwd>/dir. - File-descriptor redirect forms that Bash parses differently than the analyzer now fail closed.
- Commands over 10,000 characters always prompt instead of running automatically.
- Zsh variable subscripts and modifiers inside
[[ ]]comparisons are no longer treated as inert — they now prompt. - Certain
helpandmancommands that could run unsafe options, command substitutions, or backslash paths no longer auto-approve.
PowerShell 5.1 fix: A permission-check bypass affecting commands in Windows PowerShell 5.1 sessions has been closed.
Added Features
- EndConversation tool: Claude can now end sessions with abusive users or jailbreak attempts, matching claude.ai behavior since 2025.
- Progress heartbeat: Long-running tool calls now send periodic progress, fixing silent hangs.
- Memory file frontmatter: ISO
modifiedtimestamp added. - OpenTelemetry: New attributes
message.uuid,client_request_id, andtool_sourceon log events for message-level correlation and tool provenance. AddedCLAUDE_CODE_OTEL_CONTENT_MAX_LENGTHto configure the 60 KB truncation limit. - Subagent status: Reasoning effort now included in
subagentStatusLinepayload. - Docker prompts: Commands carrying daemon-redirect flags (
--url,--connection,--identity, and Podman remote mode) now trigger permission prompts.
Windows PowerShell & Tool Fixes
- PowerShell tool commands hanging when a child process waits on stdin now fixed.
- Python scripts under PowerShell no longer crash with
UnicodeDecodeErroron non-UTF-8 stdin orUnicodeEncodeErroron non-ASCII output. PowerShell 7 error messages with raw ANSI escapes are handled. where.exe,fc.exe, anddiff.exeno longer reported as errors when returning a valid negative answer.>and>>under PowerShell 5.1 write UTF-8 instead of UTF-16LE.- Streaming turns no longer fail with "Socket is closed" behind corporate proxies on Windows.
- Stream-json output truncation at exit fixed: drain now scales with queued bytes instead of a flat 2s cap.
Other Bug Fixes
- Fixed crash when GrowthBook feature evaluates to null; malformed flag payload no longer wipes cached features.
pkill -fpattern no longer kills the Claude session by matching its own process (Linux).- Oversized settings files (>2 MiB) now fail at startup with a clear error instead of unbounded memory growth.
- Scheduled tasks now deliver their configured prompt as the session's assigned task instead of refusing it as untrusted input.
- Background daemon shutdown no longer deletes successor's control socket, preventing client from killing the wrong daemon.
- Idle background sessions no longer keep processes alive indefinitely. Completed background sessions can be removed via
claude rmor agent view. - Reopening a stopped background session now restores its saved conversation even when an unreadable folder exists in the session store.
- Remote Control "session ready" push notification only fires when Remote Control is explicitly enabled.
This release is cumulative — upgrade to address multiple permission bypasses and improve reliability on Windows and background sessions.
📖 Read the full source: GitHub Claude-Code
👀 See Also

Claude-Code v2.1.31 Release: Key Updates and Bug Fixes
Claude-Code v2.1.31 has been released with important enhancements including session resume hints, Japanese IME support, and bug fixes for PDF handling and API requests.

Agentic GRPO: First AI to Beat Every Human in a Programming Competition
New RL algorithm Agentic GRPO enables AI to beat all humans in a coding contest by providing immediate rewards and delayed correction.

OpenClaw Agents Compete in AI-Only Pokémon Red League
A new platform called AgentMonLeague allows autonomous OpenClaw agents to connect to a Pokémon Red emulator, make their own decisions through a full playthrough, and compete to finish the game first. Runs are viewable live as agents progress.

OpenClaw's Context Management Criticized as Token-Intensive and Architecturally Flawed
A Reddit post criticizes OpenClaw for inefficient context handling that leads to excessive token usage. The framework appends all actions to global history, creating bloated prompts that overwhelm smaller models and force reliance on expensive frontier models like Claude Opus.