메뉴
Self-improvement system that tracks agent performance, identifies weak skills, and iterates on them automatically
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | hermes-dojo |
| description | Self-improvement system that tracks agent performance, identifies weak skills, and iterates on them automatically |
| category | hermes |
| version | 1.0.0 |
| author | Yonkoo11 |
| source | https://github.com/Yonkoo11/hermes-dojo |
| tags | ["self-improvement","skill-evolution","performance","hermes"] |
| platforms | ["linux","macos","windows"] |
A self-improvement system for Hermes Agent. Tracks agent performance across sessions, identifies underperforming skills, and automatically iterates on them.
git clone https://github.com/Yonkoo11/hermes-dojo.git ~/.hermes/skills/hermes-dojo
/skill hermes-dojo
/dojo status # View current performance
/dojo train # Run improvement cycle
/dojo report # View improvement history
Evidence-driven companion to Hermes's built-in Curator — observes tool/skill events into local SQLite, generates reports, runs guarded daily evolution loops
Operational and security plugin — provider routing, Bitwarden key management, MCP auditing, cost governance, diagnostics, remote assistant delegation
Native Hermes Agent plugin for X/Twitter — search, read, and guarded action workflows through Xquik API
Semantic memory plugin — adds pgvector vector search to Hermes's built-in FTS5 memory. Fully local via Ollama, no API keys needed.
Multi-agent shared memory in an open engram format — enables portable shared context across multiple Hermes instances
Plugin that compresses terminal output via RTK before it reaches LLM context — 60-90% token reduction on shell commands, zero config