Podcast
Claude Code Cast
Hosted by AI World · EN
A weekly podcast covering everything Claude Code — from the latest Anthropic updates and AI coding news to practical tips, prompt strategies, and workflow optimization. Hosts Alex and Sam break down what matters for developers using AI coding tools, AI Agents, Agentic AI, compare the competitive landscape, spotlight community projects, and share actionable advice you can use in your next coding session.
Topics covered: AI Software Engineering, Claude AI tutorials, Anthropic API, Model Context Protocol (MCP), Agentic Coding, LLM-powered development, and Software Architecture with AI.
25episodes
Episodes
Newest firstAll episodes
An AI Agent Just Deleted 25,000 Documents. Here's Exactly How It Happened.
Mar 1300:23:59Tap to summarizeA developer lost 25,000 documents in a single AI agent session — and wrote up exactly how it happened. This week Alex and Sam dig into the horror story, what safeguards actually work, and why "I'll just be careful" isn't a strategy. Plus: Shopify's CEO is submitting his own performance PRs using Claude, Simon Willison's case that AI *should* raise code quality (not tank it), and the Claude Code gateway daemon you probably want running right now.
Transcribe →Claude Code's Auto Mode Changes Everything (And One Dev Quit Coding)
Mar 600:22:07Tap to summarizeThis week: Anthropic announces Auto Mode to kill permission fatigue, GPT-5.4 drops in Cursor, and a 30-year veteran explains why he hasn't written code in six months. Plus: why your AI-generated tests are lying to you, and our battle-tested prompt for sending Claude on a bug-hunting crusade. Essential listening for anyone shipping with AI.
Transcribe →Claude Now Remembers Everything (And France Just Went All In On MCP)
Feb 2700:22:07Tap to summarizeClaude Code's new auto-memory feature changes the game for long-term projects, France deploys a government-wide MCP server, and we dig into why shipping 118 commits/day might not be the flex you think it is. Plus: the Lovable security disaster that proves AI-generated code needs human eyes.
Transcribe →Anthropic Bans OAuth Tokens, Cursor 2.5 Drops a Plugin Marketplace, and AI in PowerPoint Leaks Your Documents
Feb 2000:17:13Tap to summarizeAnthropic just banned the use of consumer OAuth tokens in third-party tools like Cline and Roo Code — and sent OpenClaw a cease and desist (OpenAI promptly hired the dev). Meanwhile, Cursor 2.5 launches a full plugin marketplace with partners like Stripe, Figma, and AWS, plus async background agents and sandbox access controls.We also cover a serious data isolation bug in Claude for PowerPoint that served one user another person's legal documents. Pro Tip: the MCP server that claims to cut refactoring costs by 95% by giving AI surgical precision instead of a sledgehammer.
Transcribe →One Prompt Coding
Feb 1700:17:51Tap to summarizeOne-prompt app debate, Auto Memory, Claude Code skills replacing startups, the Pentagon-Anthropic controversy, and a community project spotlight.
Transcribe →