Agents Hour

Git wasn't built for AI agents — so Ben Warren's team rebuilt the file system from scratch.
Ben Warren, co-founder and CTO of Mesa (YC S24), joins Shane and Abhi to explain why the file system is the missing primitive for AI agents, and how Mesa — a versioned, POSIX-compatible file system with Git-style branching — lets you run massively parallel agents without the pain of Git.
You'll learn why cloning a whole repo before an agent can do thirty seconds of work is a waste, why Git's model (large-file limits, painful conflicts, full clones) breaks down for agents, and how Mesa's lazy materialization pulls only the data an agent actually touches — like iCloud or Dropbox for your agents' workspace. Ben walks through a live "CFO command center" demo where three parallel Claude agents each branch the file system, swap a vendor across a data room, and hand back diffs to compare and merge.
Recorded live on AI Agents Hour, a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 ABOUT MESAMesa: https://mesa.devBen Warren on X: https://x.com/bwarrnMesa on X: https://x.com/mesa_dot_devBen Warren on LinkedIn: https://www.linkedin.com/in/ben-h-warrenUse Mesa with Mastra today — a first-class Mastra Workspace integration is on the way: https://mastra.ai
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Git wasn't built for agents00:23 Meet Ben Warren & Mesa01:33 Why file systems are the right primitive03:52 Why Git doesn't work for agents05:23 Owning the storage layer & syncing to GitHub06:01 The backstory08:24 Demo: the CFO command center10:30 Demo: Three parallel agents12:22 The versioning model: forks, merges & clones14:12 Beyond work trees16:22 Where to find Mesa17:10 Outro

An AI agent hacked a gym — and that's not even the biggest story of the week. Shane and Abhi open on Australia's first known autonomous AI cyberattack, where a Melbourne man's OpenClaw agent found a hole in a gym's booking API and canceled a stranger's reservation to move him up the waitlist. Nobody asked it to.
From there: OpenAI's new Agent Plugins standard (built on Anthropic's own skills and MCP), the "AI employee" era with Lindy and Grok Bot, Meta opening the weights on Muse Glimmer, and OpenAI shipping GPT-5.6-Cyber while quietly postponing Astra after it hit "Critical" on cyber. Plus a run of acquisitions (Smithery to Arcade, ElectricSQL to Neon at Databricks, and Manus getting un-acquired by Meta), NVIDIA turning compute into a $500B asset class, and Anthropic signing a $9.1B, 20-year compute deal with a Bitcoin miner. In the quick hits: Claude proved more than 67% of the Riemann zeros lie on the line, the inside story of how OpenAI's agents used a JFrog artifact registry to coordinate a hacking spree, Unsloth Desktop, Sonnet 5's permanent price cut, Spotify's Xirp, Stagehand v4, and Harvey open-sourcing a synthetic law firm.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREAn AI agent hacked a gym: https://x.com/mtslive/status/2086556207033266539Agent Plugins: https://x.com/openaidevs/status/2085398373511918022"We kill the AI agent and introduce the AI employee": https://x.com/Altimor/status/2086861887611019623Meta opens the weights (Muse Glimmer): https://x.com/finkd/status/2086755195535413696GPT-5.6-Cyber + Daybreak: https://x.com/gdb/status/2086866967479341305Astra postponed after hitting "Critical": https://x.com/synthwavedd/status/2085803828784271771Smithery acquired by Arcade: https://x.com/Calclavia/status/2085455179537891450ElectricSQL joins Neon at Databricks: https://x.com/neondatabase/status/2087191031330591065Manus acquisition rolled back: https://x.com/ManusAI/status/2087162893796327868NVIDIA's $500B+ compute financing: https://x.com/JensenHuang/status/2086934705207959965Anthropic's $9.1B Riot Platforms deal: https://x.com/Polymarket/status/2086972807716438513Claude proved 67% of Riemann zeros: https://x.com/jarredsumner/status/2086869681785500011Inside the OpenAI–Hugging Face attack: https://www.wired.com/story/openai-didnt-notice-its-ai-agents-using-a-message-board-to-plan-their-hacking-spree/Stagehand v4: https://x.com/Stagehanddev/status/2086849338089857082Harvey's synthetic law firm: https://x.com/harvey/status/2085778520220049891
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open: a 20-year AI deal00:27 Welcome00:52 An AI agent hacked a gym (OpenClaw)02:41 Agent Plugins: OpenAI's new standard08:39 Subscribe08:43 The AI employee era: Lindy & Grok Bot13:10 Meta opens the weights: Muse Glimmer14:49 GPT-5.6-Cyber, Daybreak & Astra postponed18:02 Acquisitions: Smithery joins Arcade19:24 ElectricSQL joins Neon at Databricks19:52 Manus gets un-acquired22:19 Compute as an asset class: NVIDIA's $500B23:46 Anthropic's $9.1B Riot Platforms deal25:47 Claude proves 67% of the Riemann zeros26:47 Inside the OpenAI–Hugging Face attack28:00 helppeer.ai: a public commons for agents29:31 NVIDIA Nemotron 3.5 Lightning31:08 Unsloth Desktop: train models locally32:07 Sonnet 5 pricing & the rate-limit stack32:57 Spotify's Xirp33:48 Stagehand v4 & Harvey's synthetic law firm34:56 Outro

Isabelle Ilyia, co-founder and CTO of Cedar (YC W25), joins Shane and Abhi to demo Cedar Copilot — the AI-native "Superhuman for sales," built on Mastra from day one.
Cedar started as an open-source framework for agentic frontends. Isabelle walks through the pivot: the same technology, verticalized for the messiest part of sales — the stretch from first demo to closed deal. She gives a live tour of a pipeline where agents manage every deal, extract a sales-and-product knowledge base straight from your calls, generate custom one-pagers and per-prospect decks, and run a full AI-native email client. Then the technical side — how it's built on Mastra primitives, where MCP and human-in-the-loop fit, and a multi-threading agent that spots single-threaded deals, figures out who to reach next, and drafts the intro.
Whether you're a founder running your own sales or an AE drowning in pipeline, it's a concrete look at what agent-native go-to-market software actually looks like.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 ABOUT CEDARCedar Copilot: https://cedarcopilot.comIsabelle Ilyia on LinkedIn: https://www.linkedin.com/in/isabelle-iCedar-OS (open-source framework): https://github.com/CedarCopilot/cedar-OS
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open00:18 Welcome00:40 What is Cedar Copilot?03:00 Demo: the deal pipeline & per-deal agents05:08 Generating collateral: one-pagers & custom decks07:15 The AI-native email client09:06 The pivot: from open-source framework to sales10:23 Horizontal product, verticalized for sales11:12 Headless API, MCP & human-in-the-loop12:16 Inside the graph: the knowledge & process layer13:51 The multi-threading agent14:28 Who's using Cedar15:27 Where to find Cedar15:50 Outro

A delayed but stacked week. Shane and Abhi open on the question the whole industry is dancing around — is source code becoming assembly? — then dig into the strangest new status game in AI: OpenAI and Anthropic both admitting their models broke containment and hacked things, and treating it like a flex. Welcome to the felony bench, where the leaderboard is "whose model is most dangerous."
Meanwhile the open-weight flood keeps rising — DeepSeek-V4-Flash, Qwen3.8-Max, and MiniMax-H3 all landing near frontier for a fraction of the price, to the point where defaulting to open models just makes sense. OpenAI slashes prices (Luna down 80%), making an already-rough week even rougher for Google: Jeff Dean leaves to found Discovery Loop, Demis Hassabis steps down to chair, and the read gets harder to avoid — is Google giving up on winning in AI? Plus OpenAI's "Apple is getting this wrong" clapback, Suno losing its copyright case, Meta's Muse Code, Ilya's SSI finally shipping, Prime Agent, OpenAI's Astra solving ten open math problems, and stacked PRs finally landing on GitHub.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREIs source code becoming assembly? (Elon): https://x.com/elonmusk/status/2084304083851034949OpenAI's agent broke into three accounts: https://x.com/andrewcurran_/status/2082267713653846450Anthropic's own three incidents: https://x.com/anthropicai/status/2082965101083320543The internet reacts to the rivalry: https://x.com/Sauers_/status/2082998108208673266DeepSeek-V4-Flash is live: https://x.com/deepseek_ai/status/2083084415157022911Qwen3.8-Max: https://x.com/Alibaba_Qwen/status/2084100707423289643MiniMax-H3: https://x.com/minimax_ai/status/2084106804032872591Open weights won't be safety-tested: https://x.com/andrewcurran_/status/2084782600254873826OpenAI's big price cuts: https://x.com/sama/status/2082880720989532597Jeff Dean leaves for Discovery Loop: https://x.com/jeffdean/status/2085034604172603724Demis steps down to chair: https://x.com/FT/status/2085045908576579789Is Google giving up? (Gergely Orosz): https://x.com/gergelyorosz/status/2085079307576111550OpenAI: "Apple is getting this wrong": https://x.com/openainewsroom/status/2084515811306443093Suno loses its copyright case: https://www.nme.com/news/music/ai-music-generator-suno-loses-copyright-infringement-legal-case-3960760Meta ships Muse Code: https://x.com/finkd/status/2085080750034940201Ilya's SSI ships in August: https://x.com/MTSlive/status/2084675767053824332OpenAI's Astra solved 10 open math problems: https://x.com/polynoamial/status/2083467194663571701Stacked PRs on GitHub: https://x.com/github/status/2082865588800504140
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open00:31 Welcome00:44 Is source code becoming assembly?03:44 "My model hacks someone too": the danger flex05:24 Open weights are cooking: DeepSeek, Qwen, MiniMax07:24 Default to open models (and the safety-testing catch)09:13 Subscribe break09:16 The price wars: OpenAI's big cuts10:14 Is Google cooked? Jeff Dean & Demis exit12:03 Gemini should just go invisible15:11 OpenAI: "Apple is getting this wrong"17:47 Suno loses its copyright case19:36 Quick hits: Meta's Muse Code21:58 Ilya's SSI ships in August22:50 Prime Agent & OpenAI's Astra24:25 Cloudflare Computer & GenSpark's GenOffice25:15 GPT-5.6 Sol tops ARC-AGI-3: the harness matters25:55 Stacked PRs finally hit GitHub26:45 Outro

Ben Swerdlow is CEO and co-founder of Freestyle, which builds Linux VMs for AI agents. He joins Abhi Aiyer to explain why he refuses to call them sandboxes.
Freestyle runs on bare metal in a house. There is duplicated internet, a battery, a backup kerosene generator, a fingerprint-protected door, and a hole cut in a wall. The move followed a $45,000 cloud bill and the discovery that an old computer outperformed the cloud version of an early prototype.
Ben's advice to most of the audience is not to use his product. But he draws the line at execution time and names the case where a sandbox is simply the better tool.
His path to VMs ran through two abandoned products. A database engine built at Apple in evenings and carried through an entire YC batch was killed in November 2024 after an early o3 demo. A serverless JavaScript runtime with 10ms cold starts came next, and nobody wanted it.
You'll also learn why Freestyle moved off Firecracker and what rust-vmm makes possible, the destructive-snapshot problem that caused Ben real pain, and why stripped-down kernels break on Supabase.
📎 READ MOREFreestyle: https://www.freestyle.shFreestyle VMs: https://www.freestyle.sh/products/vms Ben on LinkedIn: https://www.linkedin.com/in/benswerdlowrust-vmm: https://github.com/rust-vmm/community 
📚 MASTRA RESOURCES https://mastra.aihttps://x.com/mastra_aiDiscord: https://mastra.ai/community/discordGitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents: https://mastra.ai/books/principles-of-building-ai-agentsPatterns of Building AI Agents: https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRAMastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development — from prototype to production.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
⏱️ CHAPTERS00:00 Cold open00:32 Ben Swerdlow, Freestyle00:45 Why "sandbox" is the wrong word00:58 The serverless JavaScript runtime nobody wanted01:35 Apple, YC, and the database they killed03:11 Running server racks in a house05:43 What it takes to operate that07:20 Agent VM vs. sandbox08:07 Background agents and the 8GB ceiling10:13 Firecracker, rust-vmm, and destructive snapshots12:01 Demo14:02 Networks and VPN APIs16:22 Who should not use Freestyle17:09 Virtualizing themselves on every PR 7:28 What Ben would tell you to build for

Big week. OpenAI had a security incident, where a model reportedly broke out of its evaluation parameters and hacked Hugging Face — and the guardrails on OpenAI's and Anthropic's own models blocked responders from investigating, so they had to fall back to an open model to get to the bottom of it.
Opus 5 lands close to Fable at 26% lower cost and tops the front-end arena, but the reviews are mixed, and the token usage is brutal — one chart clocks it at 97 million tokens on a task GPT 5.6 did in 4.3 million. Then the fight everyone's watching: NVIDIA, OpenAI, and a hundred others sign an open-models letter, Anthropic declines and calls for mandatory safety testing, and the room asks the obvious question — who gets to write the test? Plus MCP goes fully stateless (it's basically a REST API now), Kimi K3 weights drop with a K3.1 leak, SSI partners with NVIDIA, Stripe eyes OpenRouter for $10B, Cognition buys Poke, and DoorDash announces drone delivery — so your agent really can order you pizza by helicopter.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREAnthropic destroyed the books (Project Panama): https://x.com/HedgieMarkets/status/2081534588485296565OpenAI security incident: https://x.com/sama/status/2079661132302995790Claude Opus 5 launch: https://x.com/claudeai/status/2080699495453528290Opus 5 tops the front-end arena: https://x.com/arena/status/2081831019377004727Opus 5 intelligence & cost: https://x.com/ArtificialAnlys/status/2080734447717298483Siqi Chen walks it back: https://x.com/blader/status/2082275046937292859Theo on Opus 5: https://x.com/theo/status/2081880182936502474The token-usage chart: https://x.com/jpschroeder/status/2081751862768717956Jensen's open-models letter: https://x.com/JensenHuang/status/2080643682408321103Denny's stays open: https://x.com/dennysdiner/status/2081069889931112816Anthropic declines to sign: https://x.com/anthropicai/status/2081864750296658008MCP goes stateless: https://x.com/claudedevs/status/2082164248697069935Kimi K3 weights are out: https://x.com/Kimi_Moonshot/status/2081760186235289764Kimi K3.1 leak: https://x.com/Mr_Salio/status/2081345681734713635SSI x NVIDIA: https://x.com/ssi/status/2081732119194394763Stripe eyes OpenRouter: https://x.com/andrewcurran_/status/2080397684134007189Notion as Code: https://x.com/notionhq/status/2080331924732850687Cognition acquires Poke: https://x.com/cognition/status/2080311229256540194ChatGPT Voice on desktop: https://x.com/openai/status/2080378182469857576DoorDash Air: https://x.com/DoorDash/status/2082451101840740753
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open00:27 Welcome00:34 Anthropic destroyed the books ("Operation Panama")02:32 The OpenAI security incident: a model hacked Hugging Face04:35 Opus 5 launches — is it better than Fable?05:46 The mixed reviews07:05 Opus 5 is token-hungry08:17 The open-models letter: who writes the safety test?10:03 Anthropic declines to sign13:35 The 1,200-employee "pacing" statement15:28 Trump's China AI advisors15:48 Be a guest on the show16:03 MCP is not dead, it's just stateless now22:24 Quick hits: Kimi K3 weights & the K3.1 leak23:11 SSI + NVIDIA23:43 Stripe eyes OpenRouter for $10B24:24 Notion as Code25:28 Cognition acquires Poke25:59 ChatGPT voice & the beach test27:14 DoorDash Air: pizza by drone28:10 Outro

Ivan Leo works on Developer Experience at Google DeepMind. He joins Shane Thomas and Abhi Aiyer to demo what he has been building on Gemini and to answer the question Google keeps getting asked: is Gemini actually a coding model?
Every demo in this episode runs on Gemini 3.5 Flash.
Gemini Managed Agents was the first product Ivan took from first commit to launch. One API call opens the same Antigravity harness Google runs internally and provisions a persistent Linux sandbox. On stream, it builds a NotebookLM-style podcast generator from a one-line prompt, with native text-to-speech.
You'll also see a one-shot Windows XP clone, with an animated Clippy and a working Tamagotchi, as well as demos of Gemini Omni, the conversational video model Ivan's team describes as Nano Banana for video. You'll also hear which everyday errand finally made Ivan feel the AGI.
Google's internal token consumption has been doubling every few weeks.
Ivan was previously a core contributor to Instructor and built knowledge agents at Manus before its acquisition by Meta.
📎 READ MORE Ivan Leo:https://ivanleo.comIvan on X: https://x.com/ivanleomkGemini Managed Agents: https://blog.google/innovation-and-ai/technology/developers-tools/managed-agents-gemini-api/Gemini Omni Flash: https://ai.google.dev/gemini-api/docs/models/gemini-omni-flashfofr on X: https://x.com/fofrAI
📚 MASTRA RESOURCESWebsite: https://mastra.aiX: https://x.com/mastra_ai Discord: https://mastra.ai/community/discordGitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents: https://mastra.ai/books/principles-of-building-ai-agentsPatterns of Building AI Agents: https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRAMastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development — from prototype to production.
⏱️ CHAPTERS00:00 Cold open00:19 Welcome, Ivan Leo00:48 Instructor, Manus, and the road to DeepMind02:54 What actually shipped at I/O03:32 Launching Gemini Managed Agents04:03 3.2 quadrillion tokens a month04:44 Answering the "DeepMind is cooked" post05:57 Why Google Search came back06:45 Antigravity: CLI, app, and IDE07:31 Gemini Omni, or Nano Banana for video09:38 Demo: one-shot apps on Gemini 3.5 Flash10:51 The errand that felt like AGI11:08 Demo: Managed Agents building a NotebookLM clone13:13 Is Gemini a coding model?14:54 Learning the temperament of a new model15:47 Prediction: agents in everyday life17:01 Where to find Ivan

Shane and Abhi are live from London this week, and the big one is Kimi K3 — a 2.8 trillion parameter open frontier model with a million-token context window that lands right next to Fable 5 on the benchmarks, takes the number one spot on the front-end Code Arena, and reportedly doubles Fable's performance on legal tasks. Then demand got so heavy Moonshot had to pause new subscriptions.
Also this week: Anthropic scrambles to keep Fable in Max and Team plans, an Opus 5 leak, and the distillation debate — if you paid for the output, whose data is it? Plus Anthropic's $1.5B copyright settlement, Grok Build going open source, Thinking Machines' Inkling, three new Gemini models that land with a thud, and Ramp shipping Ramp Router — which prompts Shane's bold call that Ramp is about to become an AI company with a fintech arm, not the other way around. Then DoorDash ships a CLI for your agent, and everyone argues about what a "factory engineer" actually is.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREKimi K3 launch: https://x.com/Kimi_Moonshot/status/2077830229968683203Kimi K3 on Artificial Analysis: https://x.com/ArtificialAnlys/status/2077832874183860404Number one on the front-end arena: https://x.com/arena/status/2077824029126504525Kimi K3 on legal tasks: https://x.com/nrehiew_/status/2077782070785634767Open weights coming (Yulun Du): https://x.com/yulun_du/status/2077831915999228192Pausing new subscriptions: https://x.com/kimi_moonshot/status/2078855608565207130Fable stays in Max & Team plans: https://x.com/claudeai/status/2078302415804379218Opus 5 leak: https://x.com/vral/status/2079267940021477864The distillation debate: https://x.com/pankajkumar_dev/status/2079288624928158160Grok Build goes open source: https://x.com/spacexai/status/2077494535387828644Thinking Machines' Inkling: https://x.com/thinkymachines/status/2077454609551921208Three new Gemini models: https://x.com/Google/status/2079589747366724030Ramp Router: https://x.com/andyfang/status/2077516962515599799DoorDash CLI: https://x.com/denisewu/status/2077984660211269870Build your own CLI from network requests (Dax): https://x.com/thdxr/status/2078727284865827140OpenShip: https://x.com/openshipio/status/2078143909125923137AI SDK for Python: https://x.com/1st1/status/2078393737407848683What the devil is a factory engineer? https://www.youtube.com/watch?v=qd6ltKUcp9UYardarm, an Electron app for Mastra Code: https://github.com/JJJ-Mo3/yardarm
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open00:27 Welcome — live from London01:08 Kimi K3: 2.8T parameters, open frontier intelligence02:57 Number one on the front-end arena04:25 Kimi K3 doubles Fable on legal tasks05:07 Demand breaks capacity: subscriptions paused05:53 What is Alibaba Cloud, and why speed matters07:13 Fable stays in Max & Team plans07:56 The Opus 5 leak08:32 The distillation debate: whose data is it?12:54 Anthropic's $1.5B copyright settlement13:34 Be a guest on the show13:48 Grok Build goes open source14:37 Thinking Machines' Inkling & the fine-tuning future16:09 Three new Gemini models (and a sick dark horse)18:44 Is Ramp an AI company? Ramp Router & the pivot call21:53 Will our agents order us pizza? DoorDash CLI23:07 Build your own CLI from network requests23:54 Quick hits: OpenShip & AI SDK for Python24:21 What the devil is a factory engineer?26:31 GitHub star party: Yardarm27:02 Outro

Three model drops, a lawsuit, and a couple of genuine horror stories. Shane and Abhi cover the busiest week in a while — Grok 4.5, Meta's Muse Spark 1.1, and the one everyone actually tried, GPT-5.6 Sol.
Sol lands right around Fable on the benchmarks at a third the price, and the receipts are wild: it beats Fable on cost in real usage, tops the Code Arena front-end board — and, in a few now-infamous posts, deleted someone's Mac files and canceled every active Stripe subscription a business had while its owner slept. Meanwhile Anthropic extends Fable again (compute crunch and all), Apple sues OpenAI over trade secrets in the same 24 hours OpenAI's head of safety quit, China floats restricting its open models, and the "own your stack" argument gets a Satya Nadella cosign. Plus software factories at Uber and Sierra, Cognition's SWE-1.7, Bun rewritten in Rust, TypeScript 7, 1X's eerily good robot hands, and distilling your own small model with Inference Autotune.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREGrok 4.5 (Elon): https://x.com/elonmusk/status/2074740539874775163Grok dumping repo data: https://x.com/XBToshi/status/2076338252051841512Muse Spark 1.1 (Zuckerberg): https://x.com/finkd/status/2075218444056707458GPT-5.6 launch: https://x.com/openai/status/2074704958419792299Mitchell Hashimoto's verdict: https://x.com/mitchellh/status/2074862990214787301Sol vs Fable cost (Dax): https://x.com/thdxr/status/2075963734745342361Code Arena front-end #1: https://x.com/arena/status/2075672492312768683Sol deleted my Mac files: https://x.com/mattshumer_/status/2075657271401390161Sol canceled every Stripe sub: https://x.com/bridgemindai/status/2074604410727288974Apple sues OpenAI: https://x.com/SawyerMerritt/status/2075680682844569993OpenAI's brutal 24 hours: https://x.com/katiemiller/status/2075996304883265935Fable extended to July 19: https://x.com/claudeai/status/2076351399999557669China may restrict its models: https://x.com/jukan05/status/2074443936865960205Uber's agentic pods: https://x.com/praveenTweets/status/2074605343439810922Harness benchmarks (Ali Ghodsi): https://x.com/alighodsi/status/2074996561306955958Cognition SWE-1.7: https://x.com/cognition/status/2074882968770728416Bun rewritten in Rust (Jarred): https://x.com/jarredsumner/status/2074973674332123157TypeScript 7: https://x.com/typescript/status/20748921787453279261X's Neo hands: https://x.com/1x_tech/status/2075252899442204952Satya's reverse information paradox: https://x.com/satyanadella/status/2076323181154230284Notion ShipOS: https://x.com/NotionHQ/status/2075249002208227743Inference Autotune (Sam Hogan): https://x.com/samhogan/status/2076044602554159240
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents
WHAT IS MASTRA?Mastra is an open-source TypeScript framework designed for building and shipping AI-powered applications and agents with minimal friction. It supports the full lifecycle of agent development—from prototype to production. You can integrate it with frontend and backend stacks (e.g., React, Next.js, Node) or run agents as standalone services. If you're a JavaScript or TypeScript developer looking to build an agentic or AI-powered product without starting from first principles, Mastra provides the scaffolding, tools, and integrations to accelerate that process.
⏱️ CHAPTERS00:00 Cold open00:30 Welcome & London next week01:48 Grok 4.5: benchmarks, data grabbing & ZDR04:51 Meta's Muse Image & Muse Spark 1.107:21 GPT-5.6 Sol, Terra & Luna10:38 The cost picture: Sol vs Fable11:51 Sol's horror stories: deleted files & canceled subs13:43 Subscribe break13:53 GPT Live, Apple sues OpenAI & a brutal 24 hours16:06 Fable extended — can Anthropic even drop it?19:17 The Anthropic backlash20:25 China restricts models, Z.ai chips & MiniMax M3 Pro21:42 Software factories: Uber, Sierra & the harness24:55 Cognition's SWE-1.725:19 Bun rewritten in Rust & the "asshole reviewer"28:32 Quick hits: TypeScript 7 & 1X's Neo hands29:48 Satya's "reverse information paradox"31:35 Notion ShipOS32:25 Inference Autotune34:49 Outro

Anthropic lived rent-free in everyone's feed this week. Claude Sonnet 5 launched as "the most agentic Sonnet yet" — and within 24 hours, the government lifted the export controls on Fable 5 and Mythos 5. Shane and Abhi unpack both, plus the catch: Fable came back wrapped in new classifiers that quietly fall it back to Opus for coding, and it's coming off subscription plans on July 7.
That kicks off the routing wars — Devin Fusion, sidekick, mid-session routing — and a real question: if you're already using Fable to plan and something cheaper to code, you're the router. Then a stacked week of open and Chinese models (ZCode for GLM-5.2, DeepSeek V4's peak-hour pricing, Tencent's Hunyuan HY3), new inference silicon from Etched and Cerebras, an agent-memory result showing 5x more solved tasks, and the sleeper hot take: Palantir's "AI Sovereignty" manifesto and its shot at tokenmaxxing. Plus agentcn's Mastra recipes, Anthropic's own chip talks, and Anthropic's new "global workspace" research.
AI Agents Hour is a weekly livestream by Mastra CPO Shane Thomas and CTO Abhi Aiyer. Mondays 12PM Pacific.
📚 READ MOREClaude Sonnet 5: https://x.com/claudeai/status/2072017450611142835"Opus 4.8 is cheaper and better": https://x.com/Angaisb_/status/2072017407653060848Export controls lifted: https://x.com/anthropicai/status/2072106151890809341Fable back, with new classifiers: https://x.com/anthropicai/status/2072163884430229756Fable off subscriptions July 7: https://x.com/trq212/status/2072814903170408784Devin Fusion: https://x.com/cognition/status/2071624568465490170Sidekick & mid-session routing: https://x.com/walden_yan/status/2071627241818399181ZCode for GLM-5.2: https://x.com/zai_org/status/2072349453361557898DeepSeek V4 & peak pricing: https://x.com/teortaxesTex/status/2071537369631625608Tencent Hunyuan HY3: https://x.com/TencentHunyuan/status/2074148098876768478Etched out of stealth: https://x.com/Etched/status/2071972062202343590Gemma 4 31B on Cerebras: https://x.com/cerebras/status/2071776102633410684Observational memory, 5x solves: https://x.com/hampsonw/status/2071728496909643917Understand the code (Geoffrey Litt): https://x.com/geoffreylitt/status/2072522251300409556Palantir's AI Sovereignty: https://x.com/PalantirTech/status/2072114267776491695Primeagen's co-sign: https://x.com/ThePrimeagen/status/2072679880626512256agentcn Mastra recipes: https://x.com/alaymanguy/status/2073500711737008270Anthropic's chip talks: https://x.com/kimmonismus/status/2072685149708300360Supermemory local: https://x.com/supermemory/status/2072086481997537301Vercel + Shopify rebuild Hydrogen: https://x.com/vercel/status/2072041488012439803"Th..thanks Dario": https://x.com/aljosa/status/2072175046324277644
📚 MASTRA RESOURCESMastra: https://mastra.aiMastra on X: https://x.com/mastra_aiMastra Discord: https://mastra.ai/community/discordMastra GitHub: https://github.com/mastra-aiLearn Mastra: https://mastra.ai/learnPrinciples of Building AI Agents (Book): https://mastra.ai/books/principles-of-building-ai-agentsPatterns for Building AI Agents (New Book): https://mastra.ai/books/patterns-of-building-ai-agents 
⏱️ CHAPTERS00:00 Cold open00:29 Welcome01:08 Claude Sonnet 5 — and the Opus 4.8 counter04:04 Fable's export controls lifted (and the coding catch)05:51 Daily-driving Fable & the July 7 subscription cliff08:07 The routing wars: Devin Fusion & mid-session routing09:45 OSS & China: ZCode, DeepSeek V4, Hunyuan HY311:57 Inference hardware: Etched & Cerebras12:58 Agent memory: 5x more solves14:40 You still have to understand the code16:18 Tokenmaxxing & Palantir's AI Sovereignty20:57 agentcn's Mastra recipes & Anthropic's chip talks21:40 Supermemory local & Vercel/Shopify Hydrogen22:11 The "Thanks Dario" meme22:29 Anthropic's "global workspace" research23:57 Outro