ai-freer/self-improving-multi-agent

最近记录的来源活动: 2026年9月12日 08:11

SkillsMP 收录数据更新: 2026年11月9日 14:22

已收集 skills: 1

GitHub 星标: 0
GitHub Forks: 0

用 AI 助手安装

复制这段 Prompt，发给你正在使用的 AI 助手。

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/ai-freer/self-improving-multi-agent.

仓库 README

仓库 README 来源：README.md

查看 README 源文件

展开完整 README

# Self-Improving Agent (Multi-Agent Variant) Based on [self-improving](https://clawhub.ai/skills/self-improving) by [ivangdavila](https://clawhub.ai/u/ivangdavila) (upstream v1.2.16). Our version has been significantly reworked: - **Multi-agent architecture**: Three agents (Lolita/Lisa/Doubao) each maintain independent hot memory - **Three-layer write model**: `hot.md` (observations) / `corrections.md` (structured error logs) / `signals.md` (external signals) - **Auto-promote mechanism**: Daily + Weekly cron jobs extract patterns from hot layer → promote to shared-rules.md - **Official memory_search integration**: OpenClaw builtin indexer owns recall and file watching - **Embedding provider handoff**: embeddings still use the custom 18790 endpoint - **Heartbeat integration**: Upstream 1.2.16 heartbeat rules and state tracking merged in ## Structure See `SKILL.md` for full architecture and trigger rules.

正在加载仓库 README…

这个仓库中的 skills

其他计算机职业

1 个职业分类 · 已分类 100%

skill

职业分类

描述

更新

self-improving-proactive-agent

其他计算机职业

三 agent 协作的自我迭代记忆系统。实时捕获纠错/反思 → 每日自动提炼 → 回写 shared 层 → 官方 memory_search 统一检索。

2026-09-12

已展示 1 / 1 个已收集 Skill。