vibehacker
News
Sierra Blog ·

Sierra and Meta propose Personal Agent Protocol, an OAuth-based standard for personal agents to sign in to businesses

Announced Oct 6 with Genesys, Instinct, Rocket, Shopify, Stripe, and Walmart, the open standard lets a personal agent start as a guest on a company's site, then sign in with read-only or write access the user picks, and finish the job through the website, MCP or OpenAPI endpoints, or the company's own agent in one OAuth session. The v0.1 spec and a reference implementation are due later in October, so nothing is implementable yet.

Read at sierra.ai

More news

View all

JetBrains releases Mellum2.1, an Apache 2.0 12B MoE trained with RL to work as a local coding sub-agent

Released Oct 8 on Hugging Face with the same 2.5B active architecture as Mellum2, the model was post trained with reinforcement learning across millions of sandboxed runs so it can explore a codebase, edit files, and check its own changes, and JetBrains says it serves almost twice as many tokens as Qwen3.5 9B under heavy load. GGUF builds for llama.cpp, Ollama, and LM Studio, plus the multi token prediction head for vLLM, are listed as coming soon…

JetBrains Blog

GitHub Copilot will run Microsoft's MAI Code 1.1 Flash locally and auto-route between your PC and the cloud

Announced Oct 7 and due by end of October in Copilot CLI, the Copilot app, and VS Code, a 53GB quantized build of the 137B parameter (6.8B active) MoE scores 70.8% on SWE Bench Verified vs 72.6% for the cloud version, with Auto mode picking local or cloud per task or you selecting it via Windows ML or any OpenAI compatible local endpoint. Microsoft's reference machine is an RTX Spark Surface Laptop Ultra with 128GB unified memory (75.5GB peak at 256K context), and local inference alone doesn't make a session offline…

Microsoft Command Line Blog

OutSystems Agent Experience goes GA, letting Claude Code, Cursor, Codex, and Kiro build on its platform

Announced Oct 8 at OutSystems World Tour Las Vegas, your coding agent shapes the app's architecture while OutSystems generates the code deterministically and enforces existing dependencies, data relationships, testing, and governance, with deploys to public cloud, private cloud, on prem, or hybrid. Early access law firm Lowenstein Sandler says it built a production litigation matter tracker with Claude Code in under three hours…

OutSystems via iTWire

Codex Cloud quietly relaunches and can now reach your private services over Tailscale

Codex lead Tibo confirmed on Oct 8 that OpenAI "silently re shipped" Codex Cloud, which is rolling out to Plus and higher and ships with Tailscale as its only VPN option, so cloud tasks can hit internal APIs and test services on your tailnet over HTTP/HTTPS (or TCP via proxy:8088 ). Tailscale says it wasn't involved and recommends a tagged, reusable, ephemeral auth key plus tight grants, since by default Codex can reach nearly everything on your tailnet…

Tailscale Blog

Spotted something we missed? Start a thread.