My Development Workflow: How I Program with AI
In this video, I walk through my real-world AI coding workflow and show how I use autonomous AI agents to plan, write, commit, and deploy production code with minimal manual intervention. Using tools like Claude Code, OpenAI Codex, Whisper, and custom agent skills, I’ve built a system where AI handles most of the development lifecycle—from implementation to GitHub commits and automated deployments on Vercel and Railway.
I cover how I structure agent configurations, sync instructions across multiple AI tools, use voice input for faster prompting, and integrate project management with Notion. You’ll also see how I give AI secure production access, enable self-improving skills, and create an efficient agentic development environment.
If you’re a developer looking to automate your workflow, ship faster, and leverage AI coding agents effectively, this video will give you practical insights and setup ideas.
Watch my previous video on OpenClaw (formerly ClawdBot): • Clawdbot: POWERFUL AI Assistant on your Phone
Topics covered:
• AI coding agents and agentic development workflows
• Claude Code and OpenAI Codex setup and configuration
• Automating GitHub commits and production deployments
• Using Whisper for voice-to-code prompting
• Syncing agent instructions across tools
• AI-driven project management with Notion
• Building self-improving AI development systems
Tools mentioned:
Claude Code, OpenAI Codex, Whisper, GitHub Actions, Vercel, Railway, Notion, OpenClaw
0:00 - Intro
0:44 - Claude Code
1:27 - Claude Skills & Model
2:53 - Codex
3:30 - Syncing Agent Configs
4:29 - Speech to Text with Whisper
5:20 - Live Demo
6:50 - Auto Commit, Push & Deploy to Production
8:18 - Project Management with Notion
9:37 - AI-Optimized Notion CLI Workflow
10:32 - What's Next: Autonomous AI Agent Teams
Company: https://overseed.ai/
X/Twitter: https://x.com/_halshin
LinkedIn: / halshin
Twitch: / halshin