01 / Build AI agent
Start with the practical AI agent tutorial.
A routing page for readers who need a first working agent path, not another abstract framework pile.
Use the framework path only if architecture blocks the first build. Otherwise ship the smallest working loop first.
Recommended first click for readers searching for how to build an AI agent or an AI agent tutorial.
This is the default start-here route if you want a beginner-friendly tutorial that gets you to a real working agent before you widen into frameworks, tooling, or scheduling.
Inside ai agent tutorial: Build Your First Real Agent Step by Step
A practical, execution-first guide to build, run, debug, and harden your first AI agent with tools, guardrails, and production checks. If you're deciding between this tutorial-first route and the architecture-first route, use /build-ai-agent as the lane map before you branch.
Choose this if your question is how to build an AI agent from scratch and you want the clearest path to a first working loop.
Architecture-first alternative, not the default start.
This is the architecture-first alternative when you need the lightweight framework, implementation boundaries, or system shape clear before the tutorial can move.
Inside How I Built a Lightweight AI Agent Framework in Python (And Battle-Tested It in One Morning)
I built MAF — a minimal AI agent framework in Python with one core loop, typed tool schemas, and JSONL traces. If the architecture-first route is useful but you still need the broader start-here map, use /build-ai-agent before you commit.
Choose this only when architecture is the blocker and you cannot make tutorial progress until the system shape is clear.
After You Choose A Start
Bring in tooling and workflow support after the first route is clear.
These notes help once the tutorial-first or framework-first choice is made. They are follow-on support, not competing starts.
Use this after the tutorial or framework choice when you need delegation, review boundaries, and a cleaner path from prompt to verified output.
A practical field guide to running coding agents safely: scope, isolation, verification, and review.
Best once your first route is chosen and you need stronger execution discipline around it.
Use this once the tutorial or framework route is clear and the next problem is routing work between coding CLIs, preserving continuity, or setting wrapper boundaries.
A practical field guide to the layer above the coding agent: when to use native CLIs, when wrappers help, and when a full harness is worth the complexity.
Best after a first route exists and orchestration, delegation boundaries, or multi-runtime control are now the blocker.
Use this once the tutorial or framework route is chosen and you need the tooling support layer around analytics, automation, and operator workflows.
I built datafast-cli and pointed an autonomous AI agent at it. 13 commands, 2 bugs found, and the 5 principles that make CLI tools genuinely useful as AI agent tools. If you're still choosing the broader first-agent path before the tooling layer, start at /build-ai-agent.
Best after the first route is clear and you are tightening the tooling around a real build.