Claude-Code v2.1.79 adds remote control, fixes subprocess hangs, and improves memory usage

✍️ OpenClawRadar📅 Published: March 19, 2026🔗 Source
Claude-Code v2.1.79 adds remote control, fixes subprocess hangs, and improves memory usage
Ad

Claude-Code v2.1.79 Release Notes

Claude-Code v2.1.79 includes authentication improvements, bug fixes for common workflow issues, and new VSCode-specific features. This release addresses several pain points developers have encountered when integrating Claude-Code into automated workflows and improves the overall stability of the tool.

New Features and Improvements

  • Authentication: Added --console flag to claude auth login for Anthropic Console (API billing) authentication
  • Configuration: Added "Show turn duration" toggle to the /config menu
  • Memory Optimization: Improved startup memory usage by ~18MB across all scenarios
  • API Reliability: Improved non-streaming API fallback with a 2-minute per-attempt timeout, preventing sessions from hanging indefinitely
  • Plugin Development: CLAUDE_CODE_PLUGIN_SEED_DIR now supports multiple seed directories separated by the platform path delimiter (: on Unix, ; on Windows)
Ad

Bug Fixes

  • Subprocess Handling: Fixed claude -p hanging when spawned as a subprocess without explicit stdin (e.g., Python subprocess.run)
  • Keyboard Interrupt: Fixed Ctrl+C not working in -p (print) mode
  • Streaming Issues: Fixed /btw returning the main agent's output instead of answering the side question when triggered during streaming
  • Voice Mode: Fixed voice mode not activating correctly on startup when voiceEnabled: true is set
  • Navigation: Fixed left/right arrow tab navigation in /permissions
  • Terminal Title: Fixed CLAUDE_CODE_DISABLE_TERMINAL_TITLE not preventing terminal title from being set on startup
  • Status Display: Fixed custom status line showing nothing when workspace trust is blocking it
  • Rate Limiting: Fixed enterprise users being unable to retry on rate limit (429) errors
  • Session Management: Fixed SessionEnd hooks not firing when using interactive /resume to switch sessions

VSCode-Specific Updates

  • Remote Control: Added /remote-control — bridge your session to claude.ai/code to continue from a browser or phone
  • Session Titles: Session tabs now get AI-generated titles based on your first message
  • UI Fixes: Fixed the thinking pill showing "Thinking" instead of "Thought for Ns" after a response completes
  • Navigation: Fixed missing session diff button when opening sessions from the left sidebar

These updates make Claude-Code more reliable for developers who integrate it into automated workflows, particularly those using Python subprocesses or dealing with API rate limits. The memory improvements and timeout handling should reduce resource usage and prevent stuck sessions in production environments.

📖 Read the full source: GitHub Claude-Code

Ad

👀 See Also

Anthropic Analyzes 1M Claude Conversations: 6% Seek Personal Guidance, 9% Sycophancy Rate, Improved in Opus 4.7
News

Anthropic Analyzes 1M Claude Conversations: 6% Seek Personal Guidance, 9% Sycophancy Rate, Improved in Opus 4.7

Analysis of 1M Claude conversations reveals 6% seek personal guidance, with relationships having highest sycophancy (25%). Opus 4.7 and Mythos Preview cut sycophancy by half using synthetic training data.

OpenClawRadar
Anthropic restricts Claude subscription usage on third-party tools like OpenClaw
News

Anthropic restricts Claude subscription usage on third-party tools like OpenClaw

Anthropic is changing its Claude subscription policy to block usage on third-party harnesses including OpenClaw, requiring separate pay-as-you-go billing for these tools starting April 4. The company is offering a one-time credit equal to monthly subscription price and pre-purchase discounts up to 30%.

OpenClawRadar
SenseNova-U1-8B-MoT: Open Source Native Multimodal Model with NEO-Unify Architecture
News

SenseNova-U1-8B-MoT: Open Source Native Multimodal Model with NEO-Unify Architecture

SenseNova released SenseNova-U1-8B-MoT, a native multimodal model that eliminates both visual encoder and VAE, using NEO-Unify architecture for unified understanding, reasoning, and generation. It excels at text-to-infographics, image editing, and interleaved text-image generation.

OpenClawRadar
ICML 2026 Desk-Rejects 2% of Papers for LLM Review Policy Violations
News

ICML 2026 Desk-Rejects 2% of Papers for LLM Review Policy Violations

ICML 2026 rejected 497 papers (~2% of submissions) after detecting 795 reviews (~1% of all reviews) where reviewers violated explicit agreements not to use LLMs. The detection method involved watermarking PDFs with hidden LLM instructions.

OpenClawRadar