The Scout Starter Kit — the 15-minute agent from Part 0, the Part 1 checkpoint (chatbot vs workflow vs agent, Scout v1), and the $0 run-it-on-your-own-machine variants

The companion to Part 0 of AI Agents: The Real-World Course. Everything you saw on screen, ready to run. INSIDE THE ZIP: (1) AGENT.PY — the exact one-screen agent from the video: two tools, the dozen-line loop every framework hides from you, and the cost line that prints the real bill of every run (the video's run: about a third of one cent). (2) THE THREE RECEIPTS — coffee, hardware, groceries — the same little world from the video, so your first run matches ours: hardware.txt, $101.99. (3) AGENT_LOCAL.PY — the $0 variant: same loop, no API key, no packages, runs entirely on your machine via Ollama. Verified end-to-end against live Ollama before this kit was packaged — including the model advice that matters: tiny 3B models flub the tool calls on this mission; the README names the models that pass. (4) THE README — the four setup commands, the API-key walkthrough (with the $5 hard-cap trick), the core exercise (add a third tool), the stretch exercise (make it find the CHEAPEST receipt without touching the tools), and the self-check answers for both. (5) NEW — THE PART 1 CHECKPOINT (part01/): chatbot.py and workflow.py — the two 'surgeries' from Part 1 — Scout v1 with the command-line mission and the window trace, the pharmacy curveball receipt that makes the workflow answer wrong with a straight face, the $0 local variant of v1, and a README with both exercises and self-check answers. Independent course from Hyperautomation Labs — not affiliated with Anthropic.

By submitting, you'll be subscribed to the Hyperautomation AI Report newsletter — one practical AI email per week. Unsubscribe anytime.

Free. No spam. Unsubscribe anytime.