跳转到主要内容
此内容尚未提供您的语言版本,正在以英文显示。

Evolving Ai Agents

技能 已验证 活跃

Provides guidance for automatically evolving and optimizing AI agents across any domain using LLM-driven evolution algorithms. Use when building self-improving agents, optimizing agent prompts and skills against benchmarks, or implementing automated agent evaluation loops.

目的

To automate the improvement of AI agents by leveraging LLM-driven evolution, making agents smarter and more performant over time.

功能

  • LLM-driven evolution of agent prompts, skills, and memory
  • File-system based workspace contract managed via Git
  • Iterative solve-observe-evolve cycles against benchmarks
  • Pluggable interfaces for agents, benchmarks, and engines
  • Built-in seed agents and benchmarks for common domains

使用场景

  • Optimizing agent prompts and skills against measurable benchmarks
  • Building self-improving agents with automated gating and rollback
  • Evolving domain-specific tool usage and procedures
  • Implementing automated agent evaluation loops

非目标

  • Building multi-agent orchestration from scratch
  • One-shot agent tasks with no iteration needed
  • RAG pipeline optimization
  • Prompt-only optimization without skill/memory evolution

安装

npx skills add Orchestra-Research/AI-Research-SKILLs

通过 npx 运行 Vercel skills CLI(skills.sh)— 需要本地安装 Node.js,以及至少一个兼容 skills 的智能体(Claude Code、Cursor、Codex 等)。前提是仓库遵循 agentskills.io 格式。

质量评分

已验证
99 /100
1 day ago 分析

信任信号

最近提交17 days ago
星标8.3k
许可证MIT
状态
查看源代码