self-evolving-skills

Search2Skill is a framework that trains LLM agents to identify capability gaps, search external sources, and distill retrieved knowledge into reusable skills using rubric-based reinforcement learning, beating baselines on expert-domain benchmarks.

The author reflects on building with Hermes Agent, highlighting its modular architecture with separate identity, memory, and skills, and poses questions about its production use and comparisons to other agent frameworks.

MIT team released a paper on self-evolving skills for Claude Code agents, achieving 71.1% pass rate, surpassing Anthropic's skill-creator by 37 points through a Generate-Test-Verify-Co-Evolve framework.

Hermes agent features self-evolving skills that save successful troubleshooting procedures as reusable skill files, automatically managed by a background Curator to merge or archive skills without data loss.