bettyguo/harness-engineer-roadmap

Harness Engineering Roadmap is an interactive, curated guide to the engineering around LLMs — the “harness”: agent loop, tool layers, context engineering, memory, retrieval, evals, observability, and production failure modes. It’s aimed at AI researchers and engineers building agents who need practical, production-ready scaffolding rather than model-centric theory. Key features: a live clickable diagram (12 areas, 84 nodes) with vetted resource panels; three pragmatic reading paths for common needs (debugging prod, scaling, learning foundations); durable vs. current tagging for links; nightly CI link checking and quarterly sweeps; static markdown fallback and contribution guidelines. Maintained by Betty Guo. Code is MIT-licensed; content CC-BY-4.0.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.