An interactive workbench for learning agent engineering. Run Python labs in your browser, inspect failing checks, and document project evidence. Explore 82 labs, 96 lessons, and capstone blueprints. Free beta. No API key required.
What became possible in your product with Astra that was not practical before?
Maker
Astra made it practical for me, as a solo builder, to develop and verify a much broader interactive learning product. Beyond writing code, it helped connect lessons, executable Python labs, project evidence, and onboarding into one experience. It also helped verify all 82 lab reference solutions and identify gaps in saving and exports. That breadth of implementation and checking would have been difficult for me to manage alone.
Report
Maker
📌
Hi Product Hunt, I'm Bhanu, the maker of Agentic Engineering Lab.
Getting an agent to work once is a useful start. Understanding why it worked, where it fails, and how to test the next change takes practice.
I built a browser workbench for developers who want to work through those questions with Python. You can edit code, run checks, investigate failures, and write down the evidence behind a project.
The free beta includes:
- 82 browser Python labs with hints and reference solutions.
- 96 lessons connecting mechanisms to worked examples.
- 32 mini-project briefs and eight capstone blueprints.
- Notes, study planning, and an editor for project evidence.
Start with the small tool-budget exercise. The starter allows another call when the budget reaches zero. Run it, fix the boundary, and explain your decision.
You can try the browser labs without an account or API key. Guest work is temporary in the current tab. ChatGPT sign-in lets you save your learning across devices. Basic Python familiarity is required, and a desktop browser works best for coding.
This is a learning beta. Passing the practice checks is not certification or proof that a system is ready for production.
I'd especially value feedback on the first exercise, the clarity of the lab feedback, and whether the evidence editor helps you explain your work. Where did you get stuck, and what would you want to investigate next?