Claude Code (Anthropic's terminal coding agent)

  • Latest version: v2.1.162 (released 3 hours ago)
  • Key updates:
- claude agents --json now includes waitingFor field showing what a waiting session is blocked on (e.g., permission prompt) - --tools: explicitly listing Grep/Glob now provides dedicated search tools on native builds with embedded search (previously ignored) - /effort command now confirms when your chosen level will persist as the default for new sessions - Clicking a slash command in the autocomplete menu fills it into your prompt instead of running immediately; press Enter to execute - Remote Control now shows as a persistent footer pill (with session link) instead of a startup message - Renamed Windsurf to Devin Desktop in /ide, /terminal-setup, and /scroll-speed menus following the editor's rebrand - Fixed silent startup hang when config directory is read-only/unwritable — now starts with in-memory config and surfaces startup errors - Fixed WebFetch permission rules not being applied to built-in preapproved domains; explicit WebFetch(domain:...) deny/ask/allow rules now take precedence over preapproved-host auto-allow - Fixed Windows permission rules for backslash and case-variant paths; Read deny rules no longer hide files from Glob/Grep results - Fixed interrupt (Esc) at start of turn being dropped in stream-json/SDK sessions - Fixed API 400 "no low surrogate in string" errors for classifier side-queries and MCP server descriptions with emoji near truncation boundary - Fixed MCP per-server timeout config values below 1000 ms being floored to 1-second watchdog; sub-1000 ms values now ignored (fallback to MCP_TOOL_TIMEOUT or default); claude mcp get annotates them - Fixed LSP tool's workspaceSymbol operation returning no results; now accepts query parameter and passes to language server - Fixed claude agents cutting live status text (tool args, replies, prompts, exec output) at 60–120 columns on wide terminals; now uses full terminal width - Fixed claude agents truncating long session names at 40 columns; name column now grows with terminal width - Fixed claude agents attach bouncing back to session list on first try after background-service restart - Fixed claude agents Ctrl+V image paste doing nothing in dispatch input and session reply box; pasting with no image now shows a hint - Fixed backgrounding a session with ← silently losing conversation when background service cannot start; session stays as failed row wakeable with Enter
  • Changelog: https://code.claude.com/docs/en/changelog
  • GitHub releases: https://github.com/anthropics/claude-code/releases

Claude API Platform Updates

  • Latest model: Claude Opus 4.8 (released May 28, 2026)
  • Key updates:
- Claude Opus 4.8 now generally available with 1M token context window by default on Claude API, Amazon Bedrock, and Vertex AI (200k on Microsoft Foundry), 128k max output tokens - Mid-conversation system messages: On Claude Opus 4.8, you can send system messages after a user turn (subject to placement rules) in the messages array, preserving prompt cache hits - stop_details field on refusal responses now publicly documented; returns category (cyber, bio, or null) and human-readable explanation - Effort parameter defaults to "high" across all surfaces (Claude Code and Messages API) for Claude Opus 4.8 - Minimum cacheable prompt length for prompt caching reduced to 1,024 tokens on Claude Opus 4.8 (lower than Claude Opus 4.7) - Adaptive thinking: Claude Opus 4.8 triggers reasoning only when a turn needs it, reducing wasted thinking tokens vs. Claude Opus 4.7 at same effort level - Claude Opus 4.8 supports high-resolution image input (up to 2576 pixels on long edge) - Task budgets, advisor tool, and computer use now support Claude Opus 4.8 - Fast mode for Claude Opus 4.8 available as research preview on Claude API only - Advisor tool now supports max_tokens parameter to cap output per call, reducing latency and cost - No longer billed for requests returning stop_reason: "refusal" without generated output
  • Deprecation: Claude Sonnet 4 and Claude Opus 4 models will be retired on June 15, 2026
  • Platform updates: Claude Managed Agents, webhooks, multiagent orchestration, and self-hosted sandboxes now available on Claude Platform on AWS (May 29, 2026)
  • Release notes: https://platform.claude.com/docs/en/release-notes/overview
  • Model deprecation notice: https://developers.make.com/white-label-documentation/release-notes/anthropic-claude-model-deprecations-on-june-15-2026

OpenAI Codex CLI

  • Latest update: GPT-5.5 model now available in Codex (June 2026)
  • Key updates (June 2026):
- Sites plugin: Now available in preview in the Codex app; create, save, deploy, and inspect websites, dashboards, internal tools, web apps, and games hosted by OpenAI; manage hosted environment variables and secrets via app sidebar; included by default in ChatGPT Business workspaces; configurable via RBAC for ChatGPT Enterprise - ChatGPT for iOS 1.2026.146 (June 2): - Optional Face ID or passcode lock for Codex - New settings screen for choosing Queue or Steer as default follow-up behavior and toggling line wrapping for code diffs - Support for connecting to Windows machines over SSH - Added support for /side <prompt> to start a side conversation with initial question - Improved follow-up prompts, Codex home screen, and viewing changed files - Fixed issues with reconnecting, archiving threads, loading tasks, and connecting to hosts - Use Codex with Amazon Bedrock (June 1): Configure Amazon Bedrock as model provider to run Codex locally with AWS-managed authentication, account controls, and billing - Terminal placement controls v26.601 (June 1): Added Default terminal location in General settings; choose whether terminal shortcut and environment actions open tabs in bottom or right panel when bottom panel enabled - Computer use and mobile access on Windows v26.527 (May 29): - Computer Use now works on Windows; Codex can operate Windows desktop apps by seeing, clicking, and typing in foreground - Remote control now supports Windows devices; start Codex work on Windows from ChatGPT mobile or Mac, check progress remotely - Profile section shows profile details, usage stats, and token activity - Added thread coordination for local projects and worktrees, including separate background threads when explicitly requested - Expanded search for past Codex app threads to include conversation content and Git branch names - Added stable identicons for background subagents across the app - Improved keyboard shortcut settings with keypress search and reset-all action - Improved Chrome context capture for Google Docs, Sheets, and Slides tabs - Appshots, goal mode, and more v26.519 (May 21): - Appshots available in Codex app on macOS; press both Command keys to send frontmost app window to Codex with screenshot and available text - Goal mode graduated from experimental to stable; available in Codex app, IDE extension, and CLI; drive Codex toward specific objective for hours or days - Remote computer use: Codex can use desktop apps after Mac locks, including remotely via Codex Mobile; includes safeguards like short-lived authorization, covered displays, relock on local input, manual-unlock fallback - Plugin sharing via marketplace sources available for ChatGPT Business; Enterprise support coming soon - Advanced in-app browser annotations for tweaking styling (font size, colors, spacing) via annotations - Browser-use improvements: faster image asset download and extraction; structured data extraction via read-only JS sandbox - Chrome extension: no longer creates tab groups when taking over tabs; uses tab icons to indicate status - Significantly improved reliability for browser use; fixed bugs on Windows, flaky availability in non-geo-blocked regions, and other performance issues - Additional updates: Auto-review documentation expanded, in-app trust review flow for hooks, Codex for Chrome extension, enterprise governance guide updates, access token creation for trusted workflows, dictation cleanup, image lightbox controls, tooltip on realtime delegation messages, MCP shortcuts, drag-and-drop reorder queued messages, warning on model downgrade, pull request status badges, worktrees setting, new MCP shortcuts, @mentions and skill mentions in inline review comments, improved MCP tool calls and Mermaid diagram error handling, stopped terminal commands no longer appear as running
  • Changelog: https://developers.openai.com/codex/changelog
  • CLI documentation: https://developers.openai.com/codex/cli

OpenCode (opencode-ai/opencode)

  • Latest version: v1.15.13 (released ~4 days ago, ~May 31, 2026)
  • Key updates:
- Core: - Bugfix: Gateway Anthropic Opus 4.7+ adaptive reasoning now keeps summarized thinking instead of returning empty thinking blocks - Improvement: Sessions can now store custom metadata through the API and SDK - Improvement: Config now loads from the opened location upward, so directory-specific settings and provider policies apply more predictably - TUI: - Bugfix: Wrapped inline tool rows now stay aligned; failed inline tools can expand error details in place - Contributors: @shantur (session metadata support)
  • Changelog: https://opencode.ai/changelog
  • GitHub releases: https://github.com/anomalyco/opencode/releases
  • Documentation: https://opencode.ai/docs/

---
AI Coding Intelligence • Leaniverse News