Claude Code v2.1.302 (Oct 16) enables fork subagents everywhere the terminal already had them (set CLAUDE_CODE_FORK_SUBAGENT=0 to opt out) and adds CLAUDE_CODE_AUTO_MODE_MODEL to pick the model behind the client-side auto mode classifier. The release also notes that standalone Claude Design at claude.ai/design closes on December 14, 2026.
As of Oct 15, FLUX 3 Image Klein ships as open weights (base for LoRA fine tuning plus a step distilled model) with native 4K output, bounding box edits, and up to 10 reference images; the INT8 build runs on 8GB RTX GPUs via ComfyUI or Diffusers. It's under the FLUX Kommunity License (free commercial use under $5M revenue) and on the BFL API as flux 3 image klein from $0.009 per image…
As of Oct 16, start an Amp runner with amp no tui desktop (or turn on Share This Mac's Display in the macOS app) and its threads get a Desktop tab in the browser or iPhone/iPad apps for viewing and controlling the runner's screen, handy for testing iOS or desktop apps the agent just built. macOS needs Apple Silicon on macOS 26+, while Linux runners spin up a separate virtual desktop so your own session isn't captured…
PostgreSQL 19 GA moves to November 19 (RC2 on Nov 12) after 8 to 10 significant features, including SQL/PGQ property graph queries, were reverted to v20 during beta. Committer Peter Eisentraut says LLM assisted code review found long tails of edge case defects faster than volunteers could triage them, and stresses the reverts weren't caused by vibe coded features…
Anthropic's Oct 15 write up says merged Claude Code PRs more than doubled from June to September while the revert share fell over 40%, with most reading moved to Claude reviewers, blocking security review, and classifier set risk tiers where low risk changes from trusted authors need no human approval. After an agent vs agent PR looped through 50+ overnight rounds, the reviewer now stops after three rounds and re reviews only the new diff…
As of Oct 16, Beam (501B MoE, 23B active, 1M context) is on Hugging Face as reflection/Beam 1.0 501B A23B under Apache 2.0, with day 0 Transformers, SGLang, and vLLM support and a free to start API at platform.reflection.ai. Reflection also shipped Mirror CLI in beta, the terminal harness Beam was trained with, which runs parallel agents in separate worktrees and supports MCP, skills, and tool call approval…
Spotted something we missed? Start a thread.
