Skip to content
HomeNewsDigestsConceptsGuidesToolbox
AboutSubscribeUA
Subscribe

AI Today Brief

The daily AI-engineering brief. Built in public. EN · UA.

XTelegramLinkedInYouTubeRSS

Follow AI Today Brief on LinkedIn for daily AI-engineering updates and the weekly “5 shifts that changed how developers work” PDF.

Explore

NewsDigestsConceptsGuides

Company

SubscribeAdvertiseAbout

Legal

Editorial policyAI disclosurePrivacyTerms

© 2026 AI Today Brief. All rights reserved.

  1. Home/
  2. News/
  3. Distillation Playbooks and Cursor Agent Architecture

Thursday, September 17, 2026

Local Models Streamline Agent Workflows

Local Models Streamline Agent Workflows. A developer's hand is shown directing a stream of data from a local, compact model into a larger, more complex agent system, symbolizing the integration of efficient local processing for enhanced age

Learn how to replace recurring API calls with cheap local models and establish disciplined Cursor agent rules.

AI-assisted · editor-reviewed·How we use AI

In this issue · 8

  1. 1
    Models & research

    Google Launches Gemini 3.8 Live and Extended Thinking Voice Models

    Google launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, introducing near real-time parallel reasoning and background tool execution without halting conversational audio. Developers can build low-latency voice agents via the Gemini API, Google AI Studio, and streaming frameworks like LiveKit and Pipecat.

    Open full story
  2. 2
    Tools & releases

    Anthropic Unifies Workspace into One Claude with Collaborative Docs and Slides

    Anthropic is merging regular chats and Cowork into a single unified workspace called 'One Claude', adding Docs and Slides in beta. Users can generate, edit, and comment on live documents and presentation slide decks directly from any conversation using existing skills and connectors.

    Open full story
  3. 3
    Agents & MCP

    Manage Kubernetes Clusters via Local Model Context Protocol Server Container

    A new practical guide demonstrates running a containerized Kubernetes Model Context Protocol (MCP) server connected to Claude Desktop. The setup mounts local kubeconfig credentials as read-only, exposing kubectl, Helm, and ArgoCD tooling directly to natural-language agent prompts.

    Open full story
  4. 4
    Tools & releases

    RPCS3 Bans Vibe Coding as PS5 Linux Lead Quits Over "Noobs Using LLMs"

    Veteran PlayStation hacker Andy 'TheFlow0' Nguyen has abandoned the PS5 Linux project, blaming the rise of AI vibe-coders in open-source spaces. The RPCS3 PlayStation 3 emulator team has announced a ban on vibe coding participants, citing code submitted by people who do not actually understand it. Nguyen says "slop kiddies" found the last remaining hypervisor bug he had been relying on and reported it to Sony for a bounty.

    Open full story
  5. 5
    Tools & releases

    Astra Burns Tokens Noticeably Quickly, So Heavy Users Fall Back to Sol

    Users report that Astra consumes tokens noticeably quickly, even on light tasks. One user now reserves Astra for very hard problems only and recommends coming back to Sol for everything else.

    Open full story
  6. 6
    Agents & MCP

    Google Home Exposes Smart Devices to Model Context Protocol Agents

    Google Home has introduced Model Context Protocol integration, allowing autonomous agents like Claude, Hermes, and Open Claw to query home telemetry and trigger automations. The interface operates alongside Gemini for Home, requiring Google Cloud setup.

    Open full story
  7. 7
    Agents & MCP

    Six Agent Harnesses Compared: Codex, Hermes Agent, and Command Code Lead on Task Success Rate

    A comparison of six agent harnesses on task success rate shows results ranging from 19 to 21 out of 29 tasks. Codex, Hermes Agent, and Command Code tied for the lead, while Claude Code followed by a single task.

    Open full story
  8. 8
    Agents & MCP

    Listing Hosted Model Context Protocol Servers Across Major Directories and Marketplaces

    Distributing hosted Model Context Protocol servers requires navigating wildly differing verification pipelines across directories. Key integration hurdles include handling unauthenticated 401 handshake checks, obtaining Glama connector badges, and managing manifest crawlers.

    Open full story

Concepts in this brief

Gemini
Browse all news

Email digest

Get the morning AI brief

One email a day — the stories that matter for engineers, founders and tech leads. Human-edited, with links to primary sources.

  • ✓120+ sources scanned daily
  • ✓Edited by a human
  • ✓1 email per day
  • ✓EN + UA

By subscribing you agree to the privacy policy.