Claude-Code v2.1.51: Security fixes, performance improvements, and new remote control feature

What's new in Claude-Code v2.1.51
Anthropics has released Claude-Code v2.1.51 with security fixes, performance improvements, and new features for developers using AI coding agents.
Key changes and fixes
- New remote control feature: Added
claude remote-controlsubcommand for external builds, enabling local environment serving for all users. - Plugin marketplace improvements: Updated default git timeout from 30s to 120s and added
CLAUDE_CODE_PLUGIN_GIT_TIMEOUT_MSenvironment variable for configuration. Added support for custom npm registries and specific version pinning when installing plugins from npm sources. - BashTool performance: Now skips login shell (
-lflag) by default when a shell snapshot is available, improving command execution performance. Previously required settingCLAUDE_BASH_NO_LOGIN=true. - Security fixes: Fixed issue where
statusLineandfileSuggestionhook commands could execute without workspace trust acceptance in interactive mode. Fixed HTTP hooks interpolating arbitrary environment variables from header values - env var interpolation now requires explicitallowedEnvVarslist in hook configuration. - Context management: Tool results larger than 50K characters are now persisted to disk (previously 100K), reducing context window usage and improving conversation longevity.
- API stability: Fixed bug where duplicate
control_responsemessages (e.g., from WebSocket reconnects) could cause API 400 errors by pushing duplicate assistant messages into conversation. - SDK improvements: Added
CLAUDE_CODE_ACCOUNT_UUID,CLAUDE_CODE_USER_EMAIL, andCLAUDE_CODE_ORGANIZATION_UUIDenvironment variables for SDK callers to provide account info synchronously, eliminating race condition where early telemetry events lacked account metadata. - Plugin compatibility: Fixed slash command autocomplete crashing when a plugin's SKILL.md description is a YAML array or other non-string type.
- Network security: HTTP hooks now routed through sandbox network proxy when sandboxing is enabled, enforcing domain allowlist. HTTP hooks not supported for SessionStart/Setup events.
- UI improvements: The
/modelpicker now shows human-readable labels (e.g., "Sonnet 4.5") instead of raw model IDs for pinned model versions, with upgrade hint when newer version available.
📖 Read the full source: GitHub Claude-Code
👀 See Also

EU forces Meta to let rival AI chatbots onto WhatsApp for free
EU regulators have ordered Meta to allow rival AI chatbots free access to WhatsApp data for integration, starting June 2026.

Apple Offers Free Private Cloud Compute to Indie Developers with Under 2M Downloads
Apple announced at WWDC 2026 that developers with fewer than 2 million first-time App Store downloads can use its Foundation Models in Private Cloud Compute with no cloud API cost. The framework also gains image input and support for server models.

Microsoft Ends Revenue Sharing with OpenAI, Impact on AI Agents Unclear
Microsoft will stop sharing revenue with OpenAI, its main AI partner, according to a Bloomberg report. The move could affect how developers integrate AI agents via Azure OpenAI services.

Harmonic-9B: Two-stage Qwen3.5-9B fine-tune for AI agents
Developer DJLougen has released Harmonic-9B, a Qwen3.5-9B fine-tune optimized for agent use with a two-stage training approach. Stage 1 (heavy reasoning) is complete, while Stage 2 (light tool-calling) is still training. GGUF quantized versions are already available.