{
    "version": "https://jsonfeed.org/version/1.1",
    "title": "AgentStackList: Trending AI tricks",
    "home_page_url": "https://agentstacklist.com/trending/",
    "feed_url": "https://agentstacklist.com/trending/feed.json",
    "items": [
        {
            "id": "nano-banana-sql-diorama-prompt",
            "url": "https://agentstacklist.com/trending/gemini/nano-banana-sql-diorama-prompt/",
            "title": "Nano Banana prompt pack: SQL-style \"planet in a hand\" city dioramas",
            "content_text": "A prompt creator shared a structured prompt for Nano Banana 2.1 that is written like a SQL query. It asks for a 2x2 grid of four cities, each shown as a tiny spherical planet resting on an open palm, with the city's main monument on top, dense blocks around it, highways wrapping the sphere, local trees, and a tilt-shift, bokeh look. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/Gdgtify/status/2111355279698805167",
            "date_published": "2026-10-17T12:00:00Z",
            "tags": [
                "gemini",
                "prompts",
                "nano-banana",
                "prompts",
                "image-generation"
            ],
            "external_url": "https://x.com/Gdgtify/status/2111355279698805167"
        },
        {
            "id": "gemini-api-cache-miss-rate",
            "url": "https://agentstacklist.com/trending/gemini/api-cache-miss-rate/",
            "title": "A developer reports high cache-miss rates on the Gemini API",
            "content_text": "Developer and creator Theo posted that Gemini is about 50% more expensive than it should be for his workload because caching isn't working well, citing a 27% cache miss rate. The post drew a lot of attention from other developers. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/theo/status/2111370893561909423",
            "date_published": "2026-10-17T12:00:00Z",
            "tags": [
                "gemini",
                "news",
                "gemini-api",
                "caching",
                "costs"
            ],
            "external_url": "https://x.com/theo/status/2111370893561909423"
        },
        {
            "id": "claude-slides-decks",
            "url": "https://agentstacklist.com/trending/claude/slides-decks/",
            "title": "Build slide decks with Claude Slides instead of hunting for templates",
            "content_text": "Boris Cherny, who leads Claude Code at Anthropic, says Claude Slides is now how he makes decks, including the device-frame mockups he used to search for by hand. Slides lives in the Claude app alongside Claude Docs and Claude Design, so you can draft a one-pager, turn it into a deck and mock up a matching visual in one conversation. Source: https://x.com/bcherny/status/2111158659971699136",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "claude",
                "writing",
                "slides",
                "presentations",
                "design"
            ],
            "external_url": "https://x.com/bcherny/status/2111158659971699136"
        },
        {
            "id": "claude-code-review",
            "url": "https://agentstacklist.com/trending/claude/code-review/",
            "title": "Let Claude Code Review check every pull request",
            "content_text": "Boris Cherny says Claude Code Review runs on all of his pull requests and catches almost every bug before it ships, and that it is part of how his team moves fast without lowering the quality bar. It is worth switching on if you merge a lot of agent-written code and want a second reviewer that never gets tired. Source: https://x.com/bcherny/status/2111158313463525518",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "claude",
                "coding",
                "claude code",
                "code review",
                "pull requests"
            ],
            "external_url": "https://x.com/bcherny/status/2111158313463525518"
        },
        {
            "id": "chatgpt-parallel-agents-pro",
            "url": "https://agentstacklist.com/trending/chatgpt/parallel-agents-pro/",
            "title": "Split one task across several agents in GPT-6 Astra Pro",
            "content_text": "OpenAI upgraded GPT-6 Astra Pro in ChatGPT on October 16 with faster responses and better use of plugins and tools. Multiple agents can now work on different parts of one task in parallel, and you can follow each agent's progress and open finished parts to review them. It is aimed at Pro users with big, splittable jobs like research across many sources. Source: https://x.com/OpenAI/status/2111192259073094035",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "chatgpt",
                "agents",
                "gpt-6 pro",
                "parallel agents",
                "research"
            ],
            "external_url": "https://x.com/OpenAI/status/2111192259073094035"
        },
        {
            "id": "codex-maintenance-backlog",
            "url": "https://agentstacklist.com/trending/chatgpt/codex-maintenance-backlog/",
            "title": "Use Codex to chip away at your maintenance backlog",
            "content_text": "OpenAI Developers shared how OpenAI's own engineers use Codex to dig through their codebase, find what needs attention and work through the maintenance backlog. It is a practical pattern for the boring work (dead code, outdated dependencies, flaky tests) that never gets prioritised. Source: https://x.com/OpenAIDevs/status/2111213093577691337",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "chatgpt",
                "coding",
                "codex",
                "maintenance",
                "refactoring"
            ],
            "external_url": "https://x.com/OpenAIDevs/status/2111213093577691337"
        },
        {
            "id": "codex-dot-on-call",
            "url": "https://agentstacklist.com/trending/chatgpt/codex-dot-on-call/",
            "title": "Put your dot on call and steer Codex from your phone",
            "content_text": "An OpenAI engineer used his dot to triage a 3am on-call incident: the dot investigated the issue and started a Codex thread, and he steered the fix from his phone in the ChatGPT app. It shows how dots and Codex hand work to each other so you do not need a laptop to start an investigation. Source: https://x.com/OpenAIDevs/status/2111211411766399288",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "chatgpt",
                "agents",
                "dots",
                "codex",
                "on-call",
                "mobile"
            ],
            "external_url": "https://x.com/OpenAIDevs/status/2111211411766399288"
        },
        {
            "id": "gpt-live-voice-design",
            "url": "https://agentstacklist.com/trending/chatgpt/gpt-live-voice-design/",
            "title": "Shape a custom voice for your app with a prompt (GPT-Live-1)",
            "content_text": "OpenAI Developers published a guide on giving GPT-Live-1 apps a voice that fits: you can shape a voice with a prompt or choose from new international preset voices. Useful if your voice agent sounds generic or wrong for your audience. Source: https://x.com/OpenAIDevs/status/2111142258867335430",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "chatgpt",
                "agents",
                "voice",
                "api",
                "developers"
            ],
            "external_url": "https://x.com/OpenAIDevs/status/2111142258867335430"
        },
        {
            "id": "gemini-canvas-pdf-to-app",
            "url": "https://agentstacklist.com/trending/gemini/canvas-pdf-to-app/",
            "title": "Turn a dense PDF into an interactive web app with Canvas",
            "content_text": "With Gemini 4 Argon, Canvas can turn a dense PDF into an interactive, animated web app you can explore at your own pace. It is a great way to make a research paper, manual or report actually readable. Source: https://x.com/GeminiApp/status/2111147479248855484",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "gemini",
                "research",
                "canvas",
                "pdf",
                "learning"
            ],
            "external_url": "https://x.com/GeminiApp/status/2111147479248855484"
        },
        {
            "id": "ai-studio-argon-pro",
            "url": "https://agentstacklist.com/trending/gemini/ai-studio-argon-pro/",
            "title": "Google AI Pro subscribers can now build with Argon in AI Studio",
            "content_text": "As of October 16, Google AI Pro subscribers can build with Gemini 4 Argon in AI Studio Build and the Playground. Before that it was limited to paid API and Ultra subscribers. It is the cheapest way to prototype small apps with Argon if you already pay for AI Pro. Source: https://x.com/GoogleAIStudio/status/2111240722426802258",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "gemini",
                "coding",
                "ai studio",
                "prototyping"
            ],
            "external_url": "https://x.com/GoogleAIStudio/status/2111240722426802258"
        },
        {
            "id": "gemini-argon-minecraft",
            "url": "https://agentstacklist.com/trending/gemini/argon-minecraft/",
            "title": "Watch four Argon agents beat Minecraft with code alone",
            "content_text": "Google DeepMind ran four Gemini 4 Argon agents in a sandbox, using only code and text. They wrote 2,783 programs, split up jobs over chat, shared notes and debugged together to beat Minecraft in 27 hours with no human help. A good read on how multi-agent teams coordinate. Source: https://x.com/GoogleDeepMind/status/2111127991401247033",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "gemini",
                "agents",
                "multi-agent",
                "research"
            ],
            "external_url": "https://x.com/GoogleDeepMind/status/2111127991401247033"
        },
        {
            "id": "cursor-cloud-agent-canvases",
            "url": "https://agentstacklist.com/trending/cursor/#cloud-agent-canvases",
            "title": "Have Cursor cloud agents build shareable canvases",
            "content_text": "Cursor cloud agents can now build interactive canvases. You can ask for a dashboard or report while the agent works, then share the finished canvas with your team as a link. Source: https://x.com/cursor_ai/status/2111193957527146711",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "cursor",
                "coding",
                "cursor",
                "cloud agents",
                "canvas"
            ],
            "external_url": "https://x.com/cursor_ai/status/2111193957527146711"
        },
        {
            "id": "claude-mythos-quantum-circuits",
            "url": "https://agentstacklist.com/trending/claude/mythos-quantum-circuits/",
            "title": "Claude Mythos 5 designs more efficient quantum circuits",
            "content_text": "From Anthropic's STEM Fellows Program: given plain-language instructions, Claude Mythos 5 designed quantum circuits that beat the best published ones, sometimes by an order of magnitude or more. More efficient circuits need fewer resources, which could bring quantum applications in chemistry and materials science closer. Source: https://x.com/AnthropicAI/status/2111152183240511640",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "claude",
                "research",
                "science",
                "quantum",
                "mythos"
            ],
            "external_url": "https://x.com/AnthropicAI/status/2111152183240511640"
        },
        {
            "id": "chatgpt-slides",
            "url": "https://agentstacklist.com/trending/chatgpt/slides/",
            "title": "ChatGPT Slides: build and co-edit decks inside ChatGPT",
            "content_text": "ChatGPT Slides lets you make a presentation from a conversation, your files or an outline, or from an example deck that sets the style. You can edit text, images, shapes and charts yourself or ask ChatGPT to change them, invite teammates to edit at the same time and leave comments, then present from ChatGPT or export to PowerPoint, PDF or Google Slides. Source: https://help.openai.com/en/articles/6825453-chatgpt-release-notes#chatgpt-slides",
            "date_published": "2026-10-16T12:00:00Z",
            "tags": [
                "chatgpt",
                "writing",
                "slides",
                "presentations",
                "space"
            ],
            "external_url": "https://help.openai.com/en/articles/6825453-chatgpt-release-notes#chatgpt-slides"
        },
        {
            "id": "chatgpt-gpt-live-voices",
            "url": "https://agentstacklist.com/trending/chatgpt/gpt-live-voices/",
            "title": "Try the new voices and accents in ChatGPT Voice (GPT-Live-1.1)",
            "content_text": "GPT-Live-1.1 rolled out to all ChatGPT users on October 15 with new voices, better accents and less \"checking...\" pausing. It adds voices in French, German, Hindi, Mexican Spanish, Korean and Japanese, and it also powers Voice in Codex and dots. OpenAI says API access is coming soon. Source: https://x.com/OpenAI/status/2110797830361436604",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "chatgpt",
                "productivity",
                "voice",
                "languages",
                "gpt-live"
            ],
            "external_url": "https://x.com/OpenAI/status/2110797830361436604"
        },
        {
            "id": "chatgpt-sites-enterprise",
            "url": "https://agentstacklist.com/trending/chatgpt/sites-enterprise/",
            "title": "Build internal team apps with ChatGPT Sites (now GA for Enterprise)",
            "content_text": "Sites became generally available for ChatGPT Enterprise on October 15, with US and EU data residency for new Sites in eligible workspaces. You can build team apps with databases and authentication, schedule updates, and let teammates use their own connected data through Plugins in Sites. Source: https://x.com/OpenAIDevs/status/2110832979656524283",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "chatgpt",
                "productivity",
                "sites",
                "enterprise",
                "internal tools"
            ],
            "external_url": "https://x.com/OpenAIDevs/status/2110832979656524283"
        },
        {
            "id": "gemini-3d-print-cad",
            "url": "https://agentstacklist.com/trending/gemini/#3d-print-cad",
            "title": "Design 3D-printable furniture add-ons from a photo",
            "content_text": "Google shows Gemini 4 Argon building an interactive 3D CAD studio in Canvas from a photo plus rough measurements. You can explore styles, adjust live sliders and export printable STL files. Source: https://x.com/GeminiApp/status/2110772859580895678",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "gemini",
                "productivity",
                "canvas",
                "3d printing",
                "cad"
            ],
            "external_url": "https://x.com/GeminiApp/status/2110772859580895678"
        },
        {
            "id": "muse-halloween-costume",
            "url": "https://agentstacklist.com/trending/muse/halloween-costume/",
            "title": "Ask your Muse for a costume idea based on your Instagram likes",
            "content_text": "Muse's own account suggested a timely trick: ask your Muse \"based on my IG likes from this year, what should I be for halloween?\" It shows off what makes Muse different, since it can draw on your connected Meta accounts for personal suggestions. Source: https://x.com/Muse/status/2110769735453561290",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "muse",
                "productivity",
                "muse",
                "instagram",
                "personal"
            ],
            "external_url": "https://x.com/Muse/status/2110769735453561290"
        },
        {
            "id": "muse-wonder-in-cursor",
            "url": "https://agentstacklist.com/trending/muse/wonder-in-cursor/",
            "title": "Use Muse Wonder inside Cursor",
            "content_text": "Cursor added Muse Wonder from Meta Superintelligence Labs on October 15. Cursor says it scores 46.1% on CursorBench at default effort, the highest of any Muse model, and it is priced at $4/M input and $20/M output tokens. You can try Meta's model without leaving your editor. Source: https://x.com/cursor_ai/status/2110867556693430293",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "muse",
                "coding",
                "cursor",
                "muse wonder"
            ],
            "external_url": "https://x.com/cursor_ai/status/2110867556693430293"
        },
        {
            "id": "muse-spark-open-weights",
            "url": "https://agentstacklist.com/trending/muse/spark-open-weights/",
            "title": "Run Meta's open-weight Muse Spark Open yourself",
            "content_text": "Meta released Muse Spark Open on October 15 and calls it the most capable US open-weight model to date: 816B total and 42B active parameters, with a 1M-token context. It is built for coding and agent work, runs on two Apple M5 Ultra Mac Studios with MLX, and is available on Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter and Together AI. Source: https://x.com/AIatMeta/status/2110779754274292212",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "muse",
                "coding",
                "open weights",
                "self-hosting"
            ],
            "external_url": "https://x.com/AIatMeta/status/2110779754274292212"
        },
        {
            "id": "chatgpt-browser-extension-tabs",
            "url": "https://agentstacklist.com/trending/chatgpt/browser-extension-tabs/",
            "title": "Let ChatGPT read your open tabs, or act in them, from the browser extension",
            "content_text": "The ChatGPT browser extension now opens a side chat beside the page you are on. It can use your open tabs, text you highlight and YouTube transcripts as context. In a Work or Codex chat on the desktop app you can @-mention Chrome, Edge, Brave, Opera or Vivaldi to have ChatGPT carry out the task in that browser, on sites where you are already signed in. By default it asks before it uses each new website. Source: https://learn.chatgpt.com/docs/chrome-extension",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "chatgpt",
                "productivity",
                "browser",
                "chrome",
                "extension",
                "work-mode"
            ],
            "external_url": "https://learn.chatgpt.com/docs/chrome-extension"
        },
        {
            "id": "muse-spark-open-on-a-mac",
            "url": "https://agentstacklist.com/trending/muse/spark-open-on-a-mac/",
            "title": "Run Meta's Muse Spark Open locally on a big Mac",
            "content_text": "Hugging Face's launch post shows how to run Muse Spark Open on Apple silicon through llama.cpp or MLX. It is a very large model, so even the 3-bit MLX quant needs more than 256 GB of RAM. Hugging Face reports about 21 tokens a second on an M3 Ultra with 512 GB. Reasoning can't be switched off, so set it to low for easy tasks to save tokens. Source: https://huggingface.co/blog/meta-muse-spark-open",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "muse",
                "coding",
                "muse-spark-open",
                "local",
                "mlx",
                "llama-cpp",
                "open-weights"
            ],
            "external_url": "https://huggingface.co/blog/meta-muse-spark-open"
        },
        {
            "id": "claude-fellowships-2027",
            "url": "https://agentstacklist.com/trending/claude/fellowships-2027/",
            "title": "Anthropic opens two new fellowships (deadline October 28)",
            "content_text": "Anthropic opened applications for two fellowships: STEM Fellows (science and engineering) and Conceptual Reasoning Fellows (philosophy, cognitive science, economics, policy and law). Fellows spend four months at Anthropic from January 2027 helping Claude reason well about hard problems in their fields. Applications close on October 28. Source: https://x.com/AnthropicAI/status/2110720940665372825",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "claude",
                "news",
                "careers",
                "fellowship",
                "anthropic"
            ],
            "external_url": "https://x.com/AnthropicAI/status/2110720940665372825"
        },
        {
            "id": "chatgpt-rosalind-academic-plan",
            "url": "https://agentstacklist.com/trending/chatgpt/rosalind-academic-plan/",
            "title": "Free ChatGPT seats for US university biology labs",
            "content_text": "OpenAI opened applications for the Rosalind Academic Plan. Approved life-science labs at US universities can get up to 30 ChatGPT Business Standard seats free for up to 12 months, with access to GPT-Rosalind subject to a trusted-access review. A principal investigator or lab lead applies for the lab. Source: https://x.com/OpenAI/status/2110841898076418336",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "chatgpt",
                "news",
                "science",
                "academic",
                "rosalind"
            ],
            "external_url": "https://x.com/OpenAI/status/2110841898076418336"
        },
        {
            "id": "synthid-time-best-inventions",
            "url": "https://agentstacklist.com/trending/gemini/synthid-time-best-inventions/",
            "title": "SynthID named one of TIME's Best Inventions of 2026",
            "content_text": "Google DeepMind's SynthID watermarking was named one of TIME's Best Inventions of 2026. Google says more than 180 billion images and videos have already been watermarked with it, which helps people tell AI-generated media from other content. Source: https://x.com/GoogleDeepMind/status/2110715601454784965",
            "date_published": "2026-10-15T12:00:00Z",
            "tags": [
                "gemini",
                "news",
                "synthid",
                "watermarking",
                "trust"
            ],
            "external_url": "https://x.com/GoogleDeepMind/status/2110715601454784965"
        },
        {
            "id": "gemini-workspace-admin",
            "url": "https://agentstacklist.com/trending/gemini/workspace-admin/",
            "title": "Turn scattered notes and photos into a deck, email and tracker",
            "content_text": "Google's example: connect your Google Workspace apps and Gemini 4 Argon turns scattered recipes, handwritten guest lists and photos of seating charts into a polished proposal deck, an email draft and an automated tracker. The pattern works for any small business drowning in admin. Source: https://x.com/GeminiApp/status/2110425806833119609",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "gemini",
                "productivity",
                "workspace",
                "small business",
                "automation"
            ],
            "external_url": "https://x.com/GeminiApp/status/2110425806833119609"
        },
        {
            "id": "claude-projects-plugin-skills",
            "url": "https://agentstacklist.com/trending/claude/projects-plugin-skills/",
            "title": "Preload your skills into every new Claude Code project",
            "content_text": "A founder shared a setup tip for Claude Code Projects: package your company or personal skills as a custom plugin and add it to your Claude web account. He says every new project you create then starts with those plugins and skills already installed, so you don't have to set them up each time. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/avthar/status/2110175403880935744",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "claude",
                "agents",
                "claude-code",
                "skills",
                "plugins"
            ],
            "external_url": "https://x.com/avthar/status/2110175403880935744"
        },
        {
            "id": "openai-eu-agents-report",
            "url": "https://agentstacklist.com/trending/chatgpt/openai-eu-agents-report/",
            "title": "OpenAI report: AI agents now drive most EU business use",
            "content_text": "OpenAI's latest economic research report says more than half of the AI use among its EU enterprise and business customers now involves agents, and the number of small business users in the region has grown 8x. The report looks at how Europe can turn broad adoption into deeper use. Source: https://x.com/OpenAI/status/2110442001749659890",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "chatgpt",
                "news",
                "europe",
                "agents",
                "research"
            ],
            "external_url": "https://x.com/OpenAI/status/2110442001749659890"
        },
        {
            "id": "ai-studio-argon-app-ideas",
            "url": "https://agentstacklist.com/trending/gemini/ai-studio-argon-app-ideas/",
            "title": "Gemini 4 Argon apps you can remix in AI Studio",
            "content_text": "Google AI Studio posted a thread of apps built with Gemini 4 Argon that you can open and try, starting with a tennis game called Argon Tennis. Each one is a working example of what Argon can build from a prompt in AI Studio Build. Source: https://x.com/GoogleAIStudio/status/2110458003027894607",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "gemini",
                "coding",
                "ai-studio",
                "argon",
                "apps"
            ],
            "external_url": "https://x.com/GoogleAIStudio/status/2110458003027894607"
        },
        {
            "id": "gemini-argon-science-breakthroughs",
            "url": "https://agentstacklist.com/trending/gemini/#argon-science-breakthroughs",
            "title": "Google says Gemini 4 Argon is helping in quantum, fusion and protein design",
            "content_text": "Google DeepMind posted a thread of science results it says Gemini 4 Argon is already helping with, including quantum computing, nuclear fusion and protein design. Source: https://x.com/GoogleDeepMind/status/2110421861435879476",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "gemini",
                "research",
                "science",
                "argon",
                "deepmind"
            ],
            "external_url": "https://x.com/GoogleDeepMind/status/2110421861435879476"
        },
        {
            "id": "gemma-scam-detection-singapore",
            "url": "https://agentstacklist.com/trending/gemini/gemma-scam-detection-singapore/",
            "title": "Google and GovTech Singapore build on-device scam detection with Gemma",
            "content_text": "Google DeepMind and GovTech Singapore are building on-device fraud detection using Google's open model Gemma. DeepMind says more than half of reported crimes in Singapore involve sophisticated social engineering. Gemma is fine-tuned on local scam tactics so it can flag threats in real time without sending personal data off the phone. Source: https://x.com/GoogleDeepMind/status/2110190101707858039",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "gemini",
                "news",
                "gemma",
                "security",
                "on-device"
            ],
            "external_url": "https://x.com/GoogleDeepMind/status/2110190101707858039"
        },
        {
            "id": "muse-health-research-papers",
            "url": "https://agentstacklist.com/trending/muse/health-research-papers/",
            "title": "Meta publishes 10 health research papers written with Muse models",
            "content_text": "Meta built a human-led research setup powered by Muse Spark 1.3 and Muse Wonder, with access to 5 million de-identified health records and public data. Medical experts used it for hypothesis generation, code and manuscript drafts while keeping control of every scientific decision. Meta shared 10 resulting manuscripts across cardiology, oncology, infectious disease, emergency medicine and more. Source: https://x.com/AIatMeta/status/2110441716507693450",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "muse",
                "research",
                "health",
                "research",
                "muse-wonder"
            ],
            "external_url": "https://x.com/AIatMeta/status/2110441716507693450"
        },
        {
            "id": "muse-wonder-agentic-benchmarks",
            "url": "https://agentstacklist.com/trending/muse/wonder-agentic-benchmarks/",
            "title": "Meta says Muse Wonder leads on agentic tasks",
            "content_text": "Meta's AI account says Muse Wonder and Muse Wonder Full reach state-of-the-art results on personal and professional agentic tasks, and links a researcher blog with the details. Meta's chief AI officer, Alexandr Wang, separately said Wonder Full topped a new coding benchmark released the same day. Source: https://x.com/AIatMeta/status/2110517571149807685",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "muse",
                "news",
                "benchmarks",
                "muse-wonder",
                "agents"
            ],
            "external_url": "https://x.com/AIatMeta/status/2110517571149807685"
        },
        {
            "id": "muse-physics-visualization",
            "url": "https://agentstacklist.com/trending/muse/physics-visualization/",
            "title": "Try Muse Wonder Full for interactive physics visualisations",
            "content_text": "Alexandr Wang, Meta's chief AI officer, shared an example of an advanced physics visualisation built with Muse Wonder Full. It's a good hint at a strength to test: turning a physics idea into an interactive, animated visual. Source: https://x.com/alexandr_wang/status/2110462535908692413",
            "date_published": "2026-10-14T12:00:00Z",
            "tags": [
                "muse",
                "research",
                "physics",
                "visualization",
                "muse-wonder-full"
            ],
            "external_url": "https://x.com/alexandr_wang/status/2110462535908692413"
        },
        {
            "id": "gemini-4-argon",
            "url": "https://agentstacklist.com/trending/gemini/4-argon/",
            "title": "Gemini 4 Argon is out: what to try first",
            "content_text": "Google's new frontier model, Gemini 4 Argon, started rolling out in the Gemini app on October 13 and is globally available. Google pitches it for hard problems from start to finish across coding, enterprise workflows and cyber defense. It is also live in the Gemini API and AI Studio for paid API and Google Ultra subscribers. Source: https://x.com/GeminiApp/status/2110039322229649783",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "research",
                "gemini 4 argon",
                "models"
            ],
            "external_url": "https://x.com/GeminiApp/status/2110039322229649783"
        },
        {
            "id": "gemini-webcam-synth",
            "url": "https://agentstacklist.com/trending/gemini/#webcam-synth",
            "title": "Play your webcam like an instrument",
            "content_text": "Gemini 4 Argon can spin up an interactive audio app in Canvas that turns your video or webcam feed into a playable synthesizer. You set the vibe and then play by moving in front of the camera. Source: https://x.com/GeminiApp/status/2110105496145576406",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "images",
                "canvas",
                "music",
                "fun"
            ],
            "external_url": "https://x.com/GeminiApp/status/2110105496145576406"
        },
        {
            "id": "cursor-gemini-argon",
            "url": "https://agentstacklist.com/trending/cursor/gemini-argon/",
            "title": "Gemini 4 Argon is now in Cursor",
            "content_text": "Cursor added Gemini 4 Argon on October 13. Cursor says it scores 52.8% on CursorBench at default effort, on par with Opus 5.5, at $2/M input and $10/M output tokens. That makes it a cheaper option for heavy agent work if it holds up on your codebase. Source: https://x.com/cursor_ai/status/2110087611763368271",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "cursor",
                "coding",
                "cursor",
                "gemini 4 argon",
                "pricing"
            ],
            "external_url": "https://x.com/cursor_ai/status/2110087611763368271"
        },
        {
            "id": "gemini-argon-verification-loop",
            "url": "https://agentstacklist.com/trending/gemini/argon-verification-loop/",
            "title": "Get better long runs from Gemini 4 Argon: give it a pass/fail check",
            "content_text": "Google's own developer guide for Gemini 4 Argon has one main tip for long autonomous tasks: give the model a way to check its own work. Point it at your test suite, logs or browser screenshots, or ask it to write a verification script, record a baseline and loop until every check passes. The guide also suggests agreeing on goals and success criteria before a long run starts, and using Argon as the orchestrator with faster Flash subagents doing the parallel work. Source: https://aistudio.google.com/learn/gemini-4-argon-developer-guide",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "coding",
                "gemini-4-argon",
                "agents",
                "testing",
                "prompting"
            ],
            "external_url": "https://aistudio.google.com/learn/gemini-4-argon-developer-guide"
        },
        {
            "id": "gemini-app-tasks-toggle",
            "url": "https://agentstacklist.com/trending/gemini/app-tasks-toggle/",
            "title": "Switch the Gemini app from chat to tasks so it keeps working while you're away",
            "content_text": "With Gemini 4 Argon, the Gemini app gained a chat-to-tasks toggle in the main window. In tasks mode Gemini can keep working on a job across your Google apps even when you are not at your computer. Google also says Gemini now connects to more than 40 apps, including Instacart, OpenTable, Adobe and Shopify. Tasks are available to Google AI Pro and Ultra subscribers. Source: https://blog.google/innovation-and-ai/products/gemini-app/gemini-4-argon-availability/",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "agents",
                "gemini-app",
                "tasks",
                "agents",
                "connectors"
            ],
            "external_url": "https://blog.google/innovation-and-ai/products/gemini-app/gemini-4-argon-availability/"
        },
        {
            "id": "claude-ambitious-prompts",
            "url": "https://agentstacklist.com/trending/claude/ambitious-prompts/",
            "title": "A four-part way to write more ambitious prompts",
            "content_text": "Felix Rieseberg, an engineering lead on Claude, published a tutorial on what \"be more ambitious with AI\" actually means. He splits it into four moves: Speed (do the same work in less time), Altitude (aim at the outcome behind the task, not just the task), Depth (ask for expert-level methods, real data and proper tools) and Range (borrow techniques from other fields). The post took off on X, with more than 20,000 likes. Source: https://x.com/felixrieseberg/status/2110023283169247623",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "claude",
                "prompts",
                "prompting",
                "claude",
                "tutorial"
            ],
            "external_url": "https://x.com/felixrieseberg/status/2110023283169247623"
        },
        {
            "id": "claude-code-lesser-known-commands",
            "url": "https://agentstacklist.com/trending/claude/code-lesser-known-commands/",
            "title": "Lesser-known Claude Code commands: /verify, /rewind and /btw",
            "content_text": "A builder shared a list of Claude Code commands he says most people miss: /verify runs your app to confirm a change really works rather than just passing tests, /code-review --fix reviews your diff and applies fixes, /rewind rolls code and conversation back to a checkpoint, and /btw asks a side question without cluttering the main session. The same post lists /goal and codex review --base main for Codex. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/Javimbuilds/status/2110074921477132532",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "claude",
                "coding",
                "claude-code",
                "commands",
                "codex"
            ],
            "external_url": "https://x.com/Javimbuilds/status/2110074921477132532"
        },
        {
            "id": "claude-check-viral-flags",
            "url": "https://agentstacklist.com/trending/claude/check-viral-flags/",
            "title": "Before you paste a viral Claude Code flag, check --help",
            "content_text": "A widely saved Claude Code tip this week told people to use a --subagents haiku flag. One developer checked and says the subagent docs list no such flag. His advice: run claude --help and grep for the option before you add anything to your config. It takes ten seconds. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/MohakBajaj5/status/2109983335388951007",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "claude",
                "coding",
                "claude-code",
                "subagents",
                "fact-check"
            ],
            "external_url": "https://x.com/MohakBajaj5/status/2109983335388951007"
        },
        {
            "id": "claude-in-house-source-control",
            "url": "https://agentstacklist.com/trending/claude/in-house-source-control/",
            "title": "Why Anthropic replaced git with its own source control",
            "content_text": "Anthropic's engineering blog explains that it went from merging about 2,500 code changes a day in May to more than 30,000 a day in September, largely because of agents. Tools built for human-scale work became the bottleneck, so the company built a new version control system and moved its monorepo off git in five months. Source: https://x.com/AnthropicAI/status/2110047081414987957",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "claude",
                "news",
                "engineering",
                "agents",
                "anthropic"
            ],
            "external_url": "https://x.com/AnthropicAI/status/2110047081414987957"
        },
        {
            "id": "gemini-argon-cornhole-robot",
            "url": "https://agentstacklist.com/trending/gemini/argon-cornhole-robot/",
            "title": "Gemini 4 Argon taught a humanoid robot to play cornhole from a video",
            "content_text": "Google DeepMind says Gemini 4 Argon watched a video of cornhole, worked out the physics, wrote its own simulation and used it to teach Apptronik's Apollo humanoid robot, which then landed seven throws in a row. Source: https://x.com/GoogleDeepMind/status/2110078384617464201",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "research",
                "robotics",
                "argon",
                "deepmind"
            ],
            "external_url": "https://x.com/GoogleDeepMind/status/2110078384617464201"
        },
        {
            "id": "awesome-nano-banana-collection",
            "url": "https://agentstacklist.com/trending/gemini/awesome-nano-banana-collection/",
            "title": "An open collection of Nano Banana edits and prompts",
            "content_text": "awesome-nano-banana is an open-source collection of Gemini Nano Banana image edits with the prompts used to make them, aimed at developers and AI artists. The post that shared it listed about 8,800 GitHub stars. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/gitrepos/status/2109932288649667047",
            "date_published": "2026-10-13T12:00:00Z",
            "tags": [
                "gemini",
                "prompts",
                "nano-banana",
                "prompts",
                "open-source"
            ],
            "external_url": "https://x.com/gitrepos/status/2109932288649667047"
        },
        {
            "id": "muse-wonder-subscribers",
            "url": "https://agentstacklist.com/trending/muse/wonder-subscribers/",
            "title": "Power your Muse with Muse Wonder (paid plans)",
            "content_text": "Meta launched Muse Wonder, its new frontier model family, on October 12. In the Muse app, Wonder powers the assistant for subscribers on the $20 Power or $100 Maximum plan, and the free tier stays on Muse Spark 1.3. Meta is rolling it out in waves, so it may not show up right away. Source: https://x.com/finkd/status/2109689305073614918",
            "date_published": "2026-10-12T12:00:00Z",
            "tags": [
                "muse",
                "agents",
                "muse wonder",
                "plans"
            ],
            "external_url": "https://x.com/finkd/status/2109689305073614918"
        },
        {
            "id": "muse-code-wonder-api",
            "url": "https://agentstacklist.com/trending/muse/code-wonder-api/",
            "title": "Code with Muse Wonder in Muse Code or the Meta Model API",
            "content_text": "For developers, Muse Wonder is available in Muse Code and the Meta Model API at $4 input and $20 output per million tokens. Muse Wonder Full is only available on Muse Code's Power and Maximum subscription plans. Meta pitches both for frontier coding, long-running agents and computer use. Source: https://dev.meta.ai/resources/blog/build-with-muse-wonder",
            "date_published": "2026-10-12T12:00:00Z",
            "tags": [
                "muse",
                "coding",
                "muse code",
                "api",
                "coding"
            ],
            "external_url": "https://dev.meta.ai/resources/blog/build-with-muse-wonder"
        },
        {
            "id": "claude-code-hooks-fail-closed",
            "url": "https://agentstacklist.com/trending/claude/code-hooks-fail-closed/",
            "title": "Make Claude Code hooks block when they break",
            "content_text": "A developer pointed out that Claude Code hooks fail open by default: if a hook script crashes, times out or has a typo in its path, you get an error notice and the command runs anyway. According to the post, version 2.1.295 added an onFailure: \"block\" setting so a broken hook stops the tool call instead. The suggested split: safety checks like lint gates, secret scans and tests should fail closed; convenience hooks like formatting and logging can fail open. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/yoshibuild/status/2109726826688028753",
            "date_published": "2026-10-12T12:00:00Z",
            "tags": [
                "claude",
                "coding",
                "claude-code",
                "hooks",
                "security"
            ],
            "external_url": "https://x.com/yoshibuild/status/2109726826688028753"
        },
        {
            "id": "claude-code-sound-cue",
            "url": "https://agentstacklist.com/trending/claude/code-sound-cue/",
            "title": "Ask your coding agent to play a sound when it needs you",
            "content_text": "A simple tip for people running several Claude Code or Codex sessions at once: add a line to your prompt asking the agent to play a sound cue when it really needs your input. You can then work on something else and only switch back when you hear it. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/ceoico/status/2109677848256508393",
            "date_published": "2026-10-12T12:00:00Z",
            "tags": [
                "claude",
                "coding",
                "claude-code",
                "codex",
                "workflow"
            ],
            "external_url": "https://x.com/ceoico/status/2109677848256508393"
        },
        {
            "id": "muse-wonder-vs-opus-game-test",
            "url": "https://agentstacklist.com/trending/muse/wonder-vs-opus-game-test/",
            "title": "Muse Wonder vs Opus 5.5: a one-prompt 3D game test",
            "content_text": "On launch day, a developer gave Muse Code (Muse Wonder, extra-high effort) and Claude Code (Opus 5.5, extra-high effort) the same one-shot job: a playable first-person 3D game of his kitchen in a single three.js file. He reports both ran first try; Opus took 11m40s and about $2.10, Wonder took 3m42s and about $1. Note: reported by users on X; we haven't independently confirmed it works in the current version. Source: https://x.com/finebutwhy/status/2109729884633534917",
            "date_published": "2026-10-12T12:00:00Z",
            "tags": [
                "muse",
                "coding",
                "muse-wonder",
                "benchmarks",
                "three-js"
            ],
            "external_url": "https://x.com/finebutwhy/status/2109729884633534917"
        }
    ]
}