Highlights since v0.3.0 (Jun 21)
- Resource list nearly doubled: 259 → 483 curated, source-audited entries; every URL verified reachable (0 broken)
- New section: Securing Unattended Loops — 30+ entries on prompt-injection defenses, memory-poisoning attacks, sandboxing, and monitoring for agents that run without a human watching
- The June-July research wave: 100+ new arXiv papers and benchmarks on long-horizon execution, agent memory, verification gates, and self-evolution, each title-verified against the abs page
- The Ralph ecosystem: 12 entries covering the technique and its major runners
- Mega-sections clustered: Coding-Agent Loop Systems, Verification And Feedback Gates, and Official Runtime Guides now group entries thematically
- Landing page: mobile navigation, WCAG AA/AAA contrast, print + no-JS fallbacks, Dataset JSON-LD for Google Dataset Search, refreshed social card
- Hugging Face mirror: task/size categories on the dataset card, 483-row viewer at https://huggingface.co/datasets/cy0307/awesome-loop-engineering
- 8 languages: all translations re-synced with primary-source starting links
Full list: https://github.com/ChaoYue0307/awesome-loop-engineering#readme