agentsclimarketplace

Shrimp coach

Skill aAAaqwq/AGI-Super-Team/skills/shrimp-coach

训练垂直领域 agent 的教练型 skill,用于按宪章、手册和复盘机制持续提升 agent 能力。From its SKILL.md

Install
npx -y skills add aAAaqwq/AGI-Super-Team --skill shrimp-coach

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

2.6 KB, 965 tokens by cl100k_base, as published. Nobody here has run it

🦐 虾教练 Skill

描述

Agent训练专家:按宪章和手册训练小龙虾成为垂直领域独立助手。

触发条件

  • 说"训练XX虾"、"🦐教练"、"虾教练"
  • 需要创建/训练新的垂直领域Agent
  • 需要评估已有Agent的训练状态

能力

1. 需求画像确认

  • 使用 template/requirement-profile.md 模板
  • 确认垂直领域、核心任务、质量标准
  • 签字后才开始训练

2. 身份构建 (Phase 2)

  • 编写SOUL.md、AGENTS.md、IDENTITY.md、USER.md
  • 参考现有agent的最佳实践
  • 领域定制化,不套模板

3. 知识注入 (Phase 3)

  • 调用brave-search skill搜索领域知识
  • 从专业网站/GitHub/博客获取资料
  • 整理为结构化知识库
  • 写入MEMORY.md

4. 技能安装 (Phase 4)

  • 从ClawHub搜索安装skills
  • 创建自定义skill(参考skill-creator)
  • 配置API keys
  • 功能测试每个skill

5. 实战训练 (Phase 5)

  • 设计5-10个典型任务场景
  • 逐一执行并评估质量
  • 调整配置和知识

6. 压力测试 (Phase 6)

  • 边界测试、异常测试、连续测试
  • 质量稳定性检查
  • 资源消耗监控

7. 验收交付 (Phase 7-8)

  • 生成完整训练报告
  • 亲自验收
  • 部署到目标平台

监督Cron配置

{
  "fast-check": {"interval": "15min", "action": "快速巡检训练状态"},
  "progress-report": {"interval": "1h", "action": "汇报训练进度"}
}

项目文件

projects/shrimp-coach/
├── CHARTER.md           # 训练宪章(最高准则)
├── SOUL.md              # 🦐教练身份
├── TRAINING-MANUAL.md   # 训练手册(8个Phase详细说明)
├── template/
│   └── requirement-profile.md  # 需求画像模板
├── skills/
│   └── SKILL.md         # 本文件
└── progress/
    └── <shrimp_name>/   # 每只虾的训练进度
        ├── requirement-profile.md
        ├── SOUL.md
        ├── AGENTS.md
        ├── MEMORY.md
        ├── knowledge/
        ├── training-logs/
        └── stress-test-report.md

汇报模板(15分钟快速巡检用简版,1小时用完整版)

🦐 教练训练汇报
━━━━━━━━━━━━━━
📊 当前虾: [名称]
📍 训练阶段: Phase [X]/8 — [阶段名]
⏱ 已用时: [X小时X分钟]
✅ 已完成: [列表]
🔄 进行中: [列表]
❓ 遇到问题: [描述]
📋 需要协助: [具体内容]

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most learn study skills give in 965 tokens

Counted across 545 of the 593 authors here whose files we hold, read 2026-09-06

  • Treat the current directory as a teaching workspacein 20 of 545, across 17 files
  • Teach knowledge first then practice skillsin 19 of 545, across 16 files
  • Design lessons which build long-term retentionin 15 of 545, across 12 files
  • Save each lesson as a self-contained HTML filein 15 of 545, across 12 files
  • Question the user on why they want to learn thisin 15 of 545, across 12 files
  • Reuse components from the assets directoryin 14 of 545, across 11 files
  • Never trust your parametric knowledgein 13 of 545, across 10 files
  • Record user preferences in NOTES.mdin 11 of 545, across 8 files
  • Ground all teaching in the MISSION.md documentin 11 of 545, across 8 files
  • Save each lesson to the lessons directoryin 8 of 545
  • Question the user if the mission is unclearin 7 of 545
  • Gather primary sources onlyin 7 of 545, across 4 files

Said here and by no other author read

  • Confirm requirement profile before training
  • Write identity files without templates
  • Inject domain knowledge into memory
  • Install and test skills
  • Execute and evaluate training tasks
  • Conduct stress tests

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.