What changed in the major AI coding CLIs?

Mem0 is now an official Anthropic connector, allowing Claude users to search and connect Mem0 directly from the Connectors directory for persistent memory across sessions [2595]. Anthropic released an official prompts library for Claude Code, providing ready-to-use prompts for development workflows [2594]. Anthropic clarifies that users can use up to half of weekly usage limit on Fable 5, then continue with credits or switch models [2593]. Claude Code's auto mode is now available without opt-in on Bedrock, Vertex AI, and Foundry [2572], alongside fixes for background sessions restoring blank [2579], auto-naming not showing in agent view [2578], malformed bracket patterns breaking file operations [2577], compound cd commands unnecessarily prompting for permission [2576], auto-updater overwriting custom launcher scripts [2575], spurious prompt-injection warnings [2574], remote managed settings bypassing consent dialogs [2573], very large markdown tables stalling rendering [2571], truncated stream-json/JSON output from `claude -p` [2570], lost replies to background agents [2569], fast mode not restoring after model switch [2568], and timeline outlines getting clipped [2591] and clipped labels and branch tooltips [2590]. Removed accidentally shipped interface transition changes [2589]. Added OAuth support for Luna Responses Lite [2588] and removed an obsolete Codex workaround interfering with OpenAI Luna Responses Lite requests [2587]. Added mouse-click support for multi-select menus in fullscreen mode [2567], `vimInsertModeRemaps` setting [2566], and screen reader mode [2565]. Added a startup warning for deprecated permission rules [2558], a live elapsed-time counter for long-running tool calls [2557], and fixed plan approvals without edits being mislabeled [2564], hook callback timeout misreported as user rejection [2563], session crash on bigint tool results [2562], paste markers leaking into external editors [2561], text fragment leaking into crash telemetry [2560], and ultracode keyword opt-in firing on non-human input [2559].[31]

Late July 2026: Mem0 becomes an official Anthropic connector [2595]. Anthropic releases an official prompts library for Claude Code [2594]. Usage limit clarification for Fable 5 [2593]. Claude Code releases a stream of reliability fixes and new features [2557–2591].[5]

OpenAI is targeting July 7 for GPT-5.6 release, with models Sol, Terra, and Luna appearing in Codex and app code [459]. Codex 5.6 is currently in testing, featuring better understanding of user intent and three new models: Sol, Terra, and Luna [985]. Codex resets have expiration dates, with the first reset expiring on July 12, and a potential global reset coinciding with GPT-5.6 release [986]. Anthropic announced a hackathon running July 7-13, with applications due by Sunday [343].[4]