Building with AI,
one honest take at a time.
We ship AI products for clients. This is where we write down what we learn — tools, patterns, mistakes, and what's actually worth your time.
Latest
How to Manage Multiple AI Agents with Omnigent
Learn how to set up Omnigent to control Claude Code, Codex, and Pi from one terminal. Manage agents, switch models, and collaborate in real time.
Turn Claude Code and Cursor Into a UI Designer With a Local Agent Skill
Install the Baoyu Design agent skill locally to generate polished UI mockups in Claude Code or Cursor, then export clean artifacts to your codebase.
Stop Giving Codex Vague Tasks: Goal Commands That Actually Work
Most Codex CLI tasks fail from vague prompts. Learn how to write structured goal commands with verification gates, constraints, and stop conditions that AI agents can deliver on.
Planner-Executor Pattern: Stop Overpaying for AI Coding
Use expensive models for architecture, cheap models for execution. Learn the planner-executor pattern with real prompt templates and cost comparisons.
Your AI Agent Works Today. Will It Work Tomorrow?
A practical framework for evaluating, testing, and monitoring AI agent workflows before you deploy them to production or deliver them to clients.
How to Give AI Agents Safe Access to Your Tools (Without Breaking Things)
A practical guide to implementing read-only scopes and granular permissions for AI agents using MCP and open-source automation tools. Keep client data safe.
How to Add MCP Servers to Claude Code
Learn how to configure MCP servers in Claude Code so your AI assistant can access project docs, APIs, and internal tools.
Build Your First MCP Server: Connect AI Agents to Your App
Step-by-step guide to building an MCP server that lets AI agents securely query your SaaS data and take real actions.
Read-Only AI Agents: Secure Automation with API Scopes
Learn how to implement read-only API authentication for AI agents using open-source workflows. Secure your automation with minimal permissions.
Build Reliable AI Agents for Client Workflows with MCP
Learn how Model Context Protocol (MCP) helps developers build secure, reliable AI agents for agency automation workflows. Practical guide for 2025.
Automate Customer Feedback Analysis with No-Code AI
Learn how to automate customer feedback analysis using no-code AI workflows. A practical guide for small business owners and agencies.
Building AI Features for Finance Apps: What Developers Need to Know
AI in finance is one of the hardest domains to build in. Regulatory boundaries, data sensitivity, and why your prompt engineering has to be different.
The Newsletter Growth Playbook for Technical Creators
How technical creators can build a useful newsletter around AI and development without hype, copied posts, or shallow automation.