🤖 OpenForgeRL: Train Harness-native Agents in Any Environment 🗓 2026-07-23 Что нового: обучение агентов end-to-end поверх реальных inference-harnessов с multi-turn reasoning и tool use, а не только в симуляторах.

🧠 Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems 🗓 2026-07-23