So I got a bunch of Tavily API credits for completing their course AND I wanted to show how to use a Boilerplate template to start apps. Combined this with a system (Claude Code Plugin) I've been developing the past couple of months I'm calling 'Apex Spec System' and I made a pretty awesome and good looking app. Complete open-source here: https://github.com/moshehbenavraham/tavily-app How it works: - Phases → major feature groups - Sessions → focused implementation units - Specs → detailed requirements per session - Task checklists → 15-30 items to complete - Validation gates → quality checks before moving on The result: - 15 sessions across 3 phases - FastAPI backend + React frontend + PostgreSQL - Auth, CRUD, 4 Tavily operations, save results with metadata - ~15K lines of production-ready code - 2 afternoons The key insight: AI doesn't drift when it has clear scope, explicit constraints, and traceable progress. It's not magic—it's just structured prompting at the project level. Video below! Curious if anyone else is experimenting with structured AI dev workflows such as BMAD, Github Spec Kit, etc. What's working for you?

Veo 3.1 is here - and it’s a major leap forward in AI video creation. Here’s what’s new: - Enhanced audio quality and realistic textures - Multi-image reference for better character and style consistency - Smooth video transitions with synced audio - Longer, continuous shots for cinematic storytelling - Powerful Insert and Remove editing tools - Available through Gemini API, Vertex AI, and the Gemini app These updates give creators more precision, flexibility, and creative control - making it easier than ever to produce high-quality, narrative-driven videos with AI. Did you test it yet?

Hey everyone! Got early access to Veo 3.1 on Higgsfield and decided to put it head-to-head with Sora 2. Did a full comparison video testing the same prompts across both platforms. Some quick takeaways: - Veo 3.1's audio quality is noticeably cleaner (almost studio-grade) - Sora 2 does longer generations (12 seconds vs 8 seconds) - Veo 3.1 works better with image-to-video when you use a starting frame - Both have their strengths depending on what you're creating I tested everything from fitness videos to anime scenes to product commercials. Also included a bunch of community-generated clips that show what's really possible with Veo 3.1. Full breakdown in video below or here: https://youtu.be/LC23Q3iw3ZM?si=ORHJiVXgr3843-ys Worth noting - Higgsfield has unlimited generations right now on Sora 2 / WAN 2.5 and almost all of their image models (such as NanoBannana and Seedream 4.0), which is huge for anyone doing production work or iterating on ideas. BONUS: I'm also giving away a comprehensive deep-research report I put together on Veo 3.1 - covers all its capabilities, technical specs, and detailed comparisons against the competition. Its available if you visit the video on YouTube Curious what you all think. Anyone else testing these tools? What's your experience been?

I just released a brand-new YouTube tutorial showing how to automate Sora 2 using N8N, Nano Banana, and Airtable — from prompt to fully rendered video. You’ll see exactly how to: - Generate images with Nano Banana - Send animation prompts to Sora 2 - Wait for renders - Update Airtable automatically with clean URLs Watch the full tutorial here: 👉 https://youtu.be/Dm4VeJnB8Qw?views The complete automation, including the n8n JSON file and setup notes, is now available in the Classroom section here. If you have questions or want me to cover multi-job execution or batch processing next, leave a comment below.

OpenAI just released big upgrades to Codex, making it a stronger coding partner: 🔹 Codex CLI – Rebuilt with community feedback. Share images (like diagrams or screenshots), track progress with to-do lists, and connect to external tools. The terminal is now cleaner and easier to follow. 🔹 IDE Extension – Codex now works inside VS Code and Cursor. It uses your open files and selections for faster, smarter help, and lets you move smoothly between cloud tasks and local edits. 🔹 Codex Cloud – Tasks now run up to 90% faster, with automatic environment setup and even browser previews of what Codex builds. 🔹 Code Review – Codex can now review GitHub pull requests, catch issues early, and even suggest or apply fixes. 💡 Codex can also work independently for up to 7 hours, handling tasks and iterating without constant supervision. I’m working on a SaaS project right now that will be available to our premium members in just a few days. Excited to share it soon!