Building self-evolving AI Agents without training
Turns pass/fail eval signals into reusable skills without retraining the model.
Self-evolving skills for your coding agent — auto-create and auto-improve skills as you work, no trigger needed.
Auto-improving agent skills without triggers, but Hermes Agent already pioneered this approach.
Developers using Claude Code, Codex, or similar AI coding agents
Hermes Agent · Cursor · Continue
inspired by hermes agent that improves itself automatically by creating skills
i've used it for a few days and i've felt the magic already - 4 skills created for me, all look reasonable.
more and better skills means faster and better job execution with fewer tokens.
supports claude code and codex now - more coming soon.
Feedback I want most: how can i make the reflection part and hooks smarter.
Turns pass/fail eval signals into reusable skills without retraining the model.
Buffer + AI video generation — promises learning, ships generic scheduling.
Recursive repair loops improve skills automatically, unlike static Claude Code defaults.
Git worktree isolation lets agents test instruction changes without breaking other sites—clever regression prevention.
Feature-packed AI agent UI, but competing against Claude Code, Cursor, and established agentic platforms.
Team-wide memory pool for agents when most tools stay siloed on one workstation.