JavaRubberDuck Podcast

0:00 Introduction

1:00 MCP: One Year in Production — What's Working, What's Broken

2:34 Prompt Injection Tops OWASP LLM Top 10 for a Third Year

4:08 Claude Code 2.1.156–159: Plugin Autoloading and Opus 4.8 Fix

5:42 Gemini CLI Retires June 18 — Migrate to Antigravity CLI Now

7:16 Ettin Reranker Family: Hugging Face Releases Six Open Rerankers

8:50 ACP Reaches Its LSP Moment: Every Major IDE and Agent on One Protocol

10:24 The AI Productivity Paradox: 93% Adoption, 10% Organisational Gain

11:58 Anthropic's $65B Raise and S-1 Filing

13:32 The OWASP Agent Top 10 and the MCP Security Gap

15:06 Claude Code June Security Sprint (2.1.160–2.1.167)

16:40 Windsurf Becomes Devin Desktop: ACP Reaches Critical Adoption

18:14 GitHub Copilot SDK Reaches GA with Rust and Java

19:48 Ollama 0.30: 20% NVIDIA Throughput Gain and Vulkan by Default

21:22 IntelliJ IDEA 2026.1.3: Terminal, Markdown, and WSL Fixes

22:56 Anthropic Files S-1 and Expands Project Glasswing

0:00 Introduction

1:00 State of Open Source AI on Hugging Face: Spring 2026

1:25 Critical RCE Vulnerabilities in AI Agent Frameworks

1:50 Securing Agentic AI: A Practical Framework for Production

2:15 How to Sandbox AI Agents: MicroVMs, gVisor, and Isolation

2:40 Microsoft 365 Copilot April 2026: Multi-Model Flexibility

3:05 OpenAI Codex 0.130.0: Plugin Sharing and Remote Control

3:30 NousResearch Hermes Agent v0.13.0

3:55 Ollama Previews MLX-Powered Inference on Apple Silicon

4:20 Anthropic and the Gates Foundation: $200M Partnership

4:45 Addy Osmani's LLM Coding Workflow: Spec First, Commit Often

5:10 Sam Altman Wraps OpenAI Trial Testimony

5:35 Cisco State of AI Security 2026

6:00 AI Agents Are Already Inside Your Perimeter

6:25 Claude Code 2.1.141-2.1.143: Plugin Enforcement, Opus 4.7

6:50 Amazon Q Developer Retiring: AWS Introduces Kiro

7:15 DeepMind AlphaEvolve: Gemini-Powered Algorithm Discovery

7:40 Anthropic Launches Claude for Small Business

8:05 Microsoft Agent Framework for .NET 1.0

8:30 IntelliJ IDEA 2026.1.2: Data-Loss Fix and MCP Path Correction

8:55 Harness Report: AI Has Outpaced Engineering Organisations

9:20 Hermes Self-Improving Agents on NVIDIA RTX and DGX Spark

9:45 Microsoft MDASH: Multi-Model Agentic Security System

10:10 Critical Ollama Vulnerabilities: Bleeding Llama Memory Leak

10:35 Hermes Agent v0.14.0: 19-Second Cold Start

11:00 xAI Launches Grok Build: Agentic CLI Coding Tool

11:25 Andrej Karpathy Joins Anthropic's Pretraining Team

11:50 Google Shuts Down Project Mariner

12:15 DeepMind Co-Scientist: Multi-Agent AI for Scientific Research

12:40 Musk vs Altman: Jury Dismisses All Claims

13:05 Google I/O 2026: Gemini 3.5 Flash and Antigravity 2.0

13:30 GitHub Copilot Removes Gemini Models from Web Interface

13:55 Software Forgets: Agent Traces Are the Memory

14:20 Docker MCP Custom Catalogs and Profiles

14:45 Coding Agent Horror Stories: The Case for Isolated Execution

15:10 Cursor Composer 2.5 Matches Frontier Benchmarks

15:35 GitHub Copilot Overhauls Plans with AI Credits

16:00 Claude Code 2.1.147-2.1.149: /code-review and Session Pinning

16:25 OpenAI Codex Appshots and Goal Mode Graduates to Standard

16:50 Docker Gordon AI Agent Reaches General Availability

17:15 NVIDIA Nemotron Diffusion LLMs: 6.4x Speedup

17:40 Karpathy at Sequoia Ascent 2026: Software 3.0

18:05 AWS Four Security Principles for Agentic AI

18:30 Claude Code 2.1.152-2.1.153: /code-review --fix

18:55 OpenAI Codex Goal Mode GA and Appshots on macOS

19:20 GitHub Copilot AI Credits Billing June 1

0:00 Introduction

1:00 METR Survey: AI Productivity Gains

3:30 Anthropic Institute for Societal Impact

6:00 AI Code Trust Gap: 96% Don't Fully Trust

8:30 Claude Code 2.1.139: Agent View & /goal

11:00 OpenAI's $4B Enterprise Deployment Company

13:30 Google DeepMind: AI Pointer