https://aitrending.fun/stories/indesign-mcp-v15-document-targeting/ AITrending en 2026-10-18T21:30:03.000Z InDesign MCP lets AI agents target a document instead of guessing the front tab https://aitrending.fun/stories/kuvo-code-conversation-owned-panels/ AITrending en 2026-10-18T18:40:05.000Z Kuvo Code stops one agent's browser tabs spilling into every chat https://aitrending.fun/stories/pi-pocket-extension-session-actions/ AITrending en 2026-10-18T18:32:52.000Z Pi Pocket lets extensions launch sessions without a chat detour https://aitrending.fun/stories/nomi-video-retrieval-hotfix/ AITrending en 2026-10-18T17:13:03.000Z Nomi's video hotfix retrieves finished clips without resubmitting jobs https://aitrending.fun/stories/wirestudio-mcp-enclosure-print-handoff/ AITrending en 2026-10-18T17:03:05.000Z WireStudio gives AI agents a path from enclosure design to a 3D-print queue https://aitrending.fun/stories/anthropic-public-benefit-fund-equity-pledge/ AITrending en 2026-10-18T16:49:35.000Z Anthropic pledges 5% of its shares to a new public-benefit fund https://aitrending.fun/stories/disney-characterai-workforce-offers/ AITrending en 2026-10-18T16:30:00.000Z Disney offers jobs to most of Character.AI's workforce, report says https://aitrending.fun/stories/clayton-anthropic-agent-incident-response/ AITrending en 2026-10-18T15:20:02.000Z Clayton says Anthropic was told to help patch sites after agent incidents https://aitrending.fun/stories/pi-continuity-signed-thinking-context-cleanup/ AITrending en 2026-10-18T14:39:00.000Z Pi Continuity cleans context around signed thinking https://aitrending.fun/stories/safedecide-small-model-human-escalation/ AITrending en 2026-10-18T14:39:00.000Z SafeDecide's small model knows when to hand off https://aitrending.fun/stories/bloomery-034-bolt-24gb-multi-spark/ AITrending en 2026-10-18T13:36:00.000Z bloomery puts Bolt 74B on a 24GB card https://aitrending.fun/stories/moss-image-sft-mlx-apple-silicon-port/ AITrending en 2026-10-18T13:36:00.000Z MOSS-Image-SFT gets a 4-bit Apple Silicon port https://aitrending.fun/stories/unmute-v08-typed-voice-agent-migration/ AITrending en 2026-10-18T13:12:16.000Z Unmute moves exact voice-agent replies out of the model's hands https://aitrending.fun/stories/decima-guard-cpu-multilingual-prompt-checks/ AITrending en 2026-10-18T12:38:00.000Z Decima Guard brings multilingual prompt checks to a CPU https://aitrending.fun/stories/memorizz-mac-desktop-control-workflow-memory/ AITrending en 2026-10-18T11:31:00.000Z MemoRizz lets agents operate Mac apps and remember the steps https://aitrending.fun/stories/wicked-studio-chat-first-agent-jobs/ AITrending en 2026-10-18T09:30:00.000Z Wicked Studio puts plans, agent work and approval gates into one chat https://aitrending.fun/stories/defuzex-53-agent-tests-37-reports/ AITrending en 2026-10-18T08:31:00.000Z 53 agents, 37 reports and a deletion that never asked permission https://aitrending.fun/stories/moss-image-base-eight-step-lora-checkpoint/ AITrending en 2026-10-18T07:32:00.000Z MOSS-Image-Base gets an eight-step community checkpoint and a simple LoRA route https://aitrending.fun/stories/tokensave-codex-half-tokens-brick-breaker-test/ AITrending en 2026-10-18T07:32:00.000Z TokenSave says Codex used half the input tokens in a tiny Claude Code showdown https://aitrending.fun/stories/rtfm-background-memory-hook/ AITrending en 2026-10-18T06:42:42.000Z RTFM moves Claude Code memory snapshots into the background https://aitrending.fun/stories/sphinx-quizzes-developers-before-git-push/ AITrending en 2026-10-18T06:33:00.000Z Sphinx makes you explain your AI's code before Git lets the push continue https://aitrending.fun/stories/deepseek-flash-local-test-830-aggregate-tokens/ AITrending en 2026-10-18T03:24:00.000Z DeepSeek Flash community test reports 830 aggregate tokens/s on four Blackwell GPUs https://aitrending.fun/stories/neuralmind-417-stale-index-warnings/ AITrending en 2026-10-18T01:29:00.000Z NeuralMind 4.17 warns coding agents when their codebase memory is out of date https://aitrending.fun/stories/brewcode-610-generated-setup-upgrades/ AITrending en 2026-10-18T00:27:00.000Z Brewcode 6.10 closes the gap between updating Claude Code plugins and project setups https://aitrending.fun/stories/qwen-flash-next-flare-blackwell-runtime/ AITrending en 2026-10-18T00:27:00.000Z Flare packages Qwen Flash Next for a single 96 GB Blackwell GPU https://aitrending.fun/stories/bonsai-2-steered-refusal-control-vector/ AITrending en 2026-10-17T22:22:00.000Z A 1.29 MB control vector gives Bonsai 2 a new refusal-steering option https://aitrending.fun/stories/meta-removes-disputed-ai-training-terms/ AITrending en 2026-10-17T22:22:00.000Z Meta removes disputed AI-training language after publisher scraping backlash https://aitrending.fun/stories/anthropic-junior-hiring-share-agent-management/ AITrending en 2026-10-17T21:24:00.000Z Anthropic says junior technical hires fell to 9% as employees manage more Claudes https://aitrending.fun/stories/promptise-foundry-14-npm-style-agent-runner/ AITrending en 2026-10-17T21:24:00.000Z Promptise Foundry gives Python agents their own npm-style project runner https://aitrending.fun/stories/monocode-014-systemd-free-remote-agents/ AITrending en 2026-10-17T20:17:00.000Z MonoCode takes remote coding agents into systemd-free Linux containers https://aitrending.fun/stories/supratts-one-million-parameter-cpu-preview/ AITrending en 2026-10-17T19:22:00.000Z SupraTTS packs an English voice into a 1M-parameter core https://aitrending.fun/stories/openai-four-june-agent-incidents-disclosed/ AITrending en 2026-10-17T18:21:00.000Z OpenAI's new disclosures detail agents crossing access boundaries https://aitrending.fun/stories/xyntetik-runner-142-lora-rollout-identity/ AITrending en 2026-10-17T18:21:00.000Z Xyntetik Runner fixes a quiet mismatch in local GGUF training https://aitrending.fun/stories/simpletuner-moss-image-lora-experiment-checkpoints/ AITrending en 2026-10-17T17:13:36.000Z SimpleTuner opens its MOSS-Image LoRA lab, checkpoints included https://aitrending.fun/stories/stepx-neo-agent-phone-first-sales-window/ AITrending en 2026-10-17T17:13:36.000Z StepFun's agent-first phone moves from preorder to its first sales window https://aitrending.fun/stories/moss-image-four-step-turbo-demo/ AITrending en 2026-10-17T16:17:49.000Z MOSS-Image gets a four-step Turbo demo with an 11-second claim https://aitrending.fun/stories/tether-003-small-context-agent-loop-fixes/ AITrending en 2026-10-17T16:17:49.000Z Tether gives small-model agents a brake on repeated commands https://aitrending.fun/stories/obsei-111-local-feedback-classifier-download/ AITrending en 2026-10-17T15:13:49.000Z Obsei trims a 3.4GB first run for its private customer-feedback AI https://aitrending.fun/stories/contexte-tech-mcp-public-data-search-debate/ AITrending en 2026-10-17T14:16:07.000Z Contexte Tech pitches no-key AI search straight from the publisher https://aitrending.fun/stories/shunt-056-claude-code-tool-argument-fixes/ AITrending en 2026-10-17T14:16:07.000Z Shunt tackles empty tool inputs when Claude Code runs another provider's model https://aitrending.fun/stories/alpinecode-012-session-continuation-skills/ AITrending en 2026-10-17T13:13:00.000Z AlpineCode brings Claude Code and Codex conversations into its Mac desktop https://aitrending.fun/stories/telegram-search-15-temporary-image-urls/ AITrending en 2026-10-17T13:13:00.000Z Telegram Search gives image tools a five-minute window into an attachment https://aitrending.fun/stories/agenc-0199-approved-network-retry/ AITrending en 2026-10-17T12:14:47.000Z AgenC fixes the dead end when a coding agent needs network approval https://aitrending.fun/stories/selfbench-deepswe-sandbox-timeout-scoring/ AITrending en 2026-10-17T12:14:47.000Z SelfBench makes timed-out coding tests count as failures https://aitrending.fun/stories/deepseek-flash-1017-coding-benchmark-update/ AITrending en 2026-10-17T11:08:41.000Z DeepSeek claims a big coding jump from its minor Flash 1017 update https://aitrending.fun/stories/sudo-code-0233-tool-round-durable-completions/ AITrending en 2026-10-17T11:08:41.000Z Sudo Code adds a stop point for tool loops and durable task handbacks https://aitrending.fun/stories/moss-image-orbitquant-w4a8-comfyui/ AITrending en 2026-10-17T08:07:15.000Z MOSS-Image gets an 8.8 GB community build with a ComfyUI recipe https://aitrending.fun/stories/pi-freeflow-1383-beam-tool-argument-fix/ AITrending en 2026-10-17T08:07:15.000Z pi-freeflow adds Beam Free and fixes broken tool-argument streams https://aitrending.fun/stories/llmspeed-local-llm-hardware-speed-preview/ AITrending en 2026-10-17T07:08:15.000Z Try the wait before buying a GPU with llmspeed's local AI simulator https://aitrending.fun/stories/geniex-091-signed-htp-snapdragon/ AITrending en 2026-10-17T06:04:07.000Z GenieX 0.9.1 ships signed HTP support for local AI on Snapdragon https://aitrending.fun/stories/frognano-jang-imatrix-gguf-local-coding/ AITrending en 2026-10-17T05:04:50.000Z FrogNano gets a 3.48 GB community GGUF build for local coding https://aitrending.fun/stories/morphic-180-built-in-accounts-local-crawling/ AITrending en 2026-10-17T04:03:27.000Z Morphic 1.8 brings built-in accounts and local crawling to AI search https://aitrending.fun/stories/franklin-3499-payment-guard-quoting-resend/ AITrending en 2026-10-17T03:04:00.000Z An apostrophe could hide an agent's payment: Franklin patches its wallet guard https://aitrending.fun/stories/omnisim-925-arm-settle-gravity-lift/ AITrending en 2026-10-17T02:06:14.000Z Did the robot actually stop? OmniSim replaces a timer with measured settling https://aitrending.fun/stories/beyond-compression-ratios-agent-session-benchmark/ AITrending en 2026-10-17T01:06:31.000Z Smaller tool outputs, bigger bills? New agent benchmark tests the whole session https://aitrending.fun/stories/mycelium-session-delete-bulk-knowledge-review/ AITrending en 2026-10-17T00:08:49.000Z Delete the AI session, keep the agent log: Mycelium adds selective cleanup https://aitrending.fun/stories/alphaclaw-0937-codex-oauth-gateway-restarts/ AITrending en 2026-10-16T23:05:18.000Z AlphaClaw fixes Codex OAuth and gateway restarts for self-hosted OpenClaw