Fetching from the wire…
Public story · 2026-03-15 · source-backed
Anthropic, Vercel, Mastra, LangGraph, and OpenAI all shipped harness primitives simultaneously — external scaffolding managing persistent state, retry logic, and HITL checkpoints around stateless LLM inference. This isn't framework preference; it's architectural necessity. The mismatch between stateless inference and stateful tasks cannot be solved at the model layer. Phil Schmid
Each link below shares sources, entities, or timing with this story.
Kits like TanStack/ai (type-safe, provider-agnostic streaming, tools, and agents across OpenAI/Anthropic/Gemini) and manifest let you swap providers, and increasingly the harness itself, without rewriting (TanStack). With the model market churning weekly and June releases slip...
The ARC-AGI-2 breakthroughs reveal a concrete architectural pattern. Symbolica's Agentica achieves 85.28% (with Opus 4.6) using recursive delegation where sub-agents spawn sub-agents, each receiving only relevant state and avoiding context rot. Average 2.6 agents per task, max...
Guillermo Rauch told TechCrunch that the industry is decoupling models from agents, with customers moving to plug-and-play stacks spanning OpenAI, Anthropic, Gemini, DeepSeek, and GLM 5.2 rather than betting on one lab. Over 1 trillion tokens daily through Vercel's AI gateway....
Version bumps are usually boring. This one isn't, because of what it concedes. The OpenAI Agents SDK hit v0.13 with an any-LLM adapter, breaking the assumption that you'd be running OpenAI models inside OpenAI's framework. Requesty's SDK comparison has the details: alongside t...
Three frontier models shipped in a single week this month, and teams with a standing eval harness had a routing decision in hours. Anthropic's own agent-eval guidance says 20-50 tasks drawn from your real usage and real failures is enough to detect issues (DeepEval). DeepEval...
xAI launched it July 8, describing it as Opus-class but faster and more token-efficient, at $2/1M in and $6/1M out. Trained across tens of thousands of NVIDIA GB300 GPUs with RL over hundreds of thousands of multi-step software engineering tasks, and trained *alongside Cursor*...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.