pilot

Pilot - Skill Navigator

Prerequisite Protocol

Environment Detection

PROJECT_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || echo "unknown")
BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
COMMIT=$(git rev-parse --short HEAD 2>/dev/null || echo "unknown")

echo "PROJECT: $PROJECT_ROOT"
echo "BRANCH: $BRANCH"
echo "COMMIT: $COMMIT"

Artifact Directory Initialization

mkdir -p memory/artifacts/pilot

Experience Retrieval (Recommended)

Trigger Timing: Before analyzing the task type, prepare best practices for the current task

Recommended Action: Call the experience-manager skill to retrieve relevant experience

Execution Steps：

Extract task keywords

从用户请求中提取：
- 核心任务关键词
- 技术栈关键词
- 问题类型关键词

Write request artifact

Write(
  file_path="memory/artifacts/pilot/experience-request.json",
  content={
    "requesting_skill": "pilot",
    "action": "retrieve",
    "task_keywords": ["关键词1", "关键词2"],
    "task_type": "bugfix|feature|refactor",
    "technologies": ["Technology1", "Technology2"]
  }
)

Read experience results

result = Read("memory/artifacts/experience-manager/experience-result.json")

Apply to skill recommendations

基于历史经验，推荐以下技能链：

1. /investigate - 快速定位问题（历史成功率 90%）
2. /first-principles - 分析根本原因（历史应用 5 次）
3. /pdca-cycle - 迭代优化

⚠️ 预防措施（来自历史经验）：
- 新项目优先检查国际化配置
- 使用标准配置模板

Value: Optimize skill recommendations based on historical experience to improve the success rate.

Error Prevention (Automatic)

Trigger Timing: When a potential error risk is detected

Automatic Action: Call experience-manager to obtain prevention suggestions

Execution Steps：

Detect risk

在分析任务类型时，检测常见的错误风险：
- 缺少配置检查
- 使用了已废弃的API
- 忽略了边界情况

Write prevention request

Write(
  file_path="memory/artifacts/pilot/error-prevention-request.json",
  content={
    "requesting_skill": "pilot",
    "action": "prevent",
    "current_step": "任务分析",
    "context": {
      "technology": "...",
      "feature": "...",
      "risk_factors": [...]
    }
  }
)

Display prevention suggestions

## 💡 错误预防提醒

检测到潜在风险：国际化配置缺失

**预防措施**：
- 优先检查 next.config.js 中的 i18n 配置
- 使用标准配置模板

**历史错误**：2026-03-20 类似错误

Value: Prevent errors in advance and avoid repeating the same pitfalls.

Overview

Intelligently analyze user requests, recommend the best methodology skill combinations, and achieve collaborative skill work.

Core Value：

Automatically identify task types
Recommend optimal skill chains
Provide execution context
Reduce decision burden

Core Principle: Let skills work collaboratively to achieve a 1+1>2 effect.

When to Use

Applicable Scenarios：

The user initiates any task request
The user is unsure which skill to use
The user wants skill recommendations
A complex task requires multiple skills to work together

Inapplicable Scenarios：

The user has explicitly specified the skill to use
Simple Q&A or information queries

The Process

digraph orchestrator {
    rankdir=TB;

    "用户请求" [shape=box, style=filled, fillcolor="#c8e6c9"];
    "分析任务类型" [shape=diamond, style=filled, fillcolor="#bbdefb"];
    "匹配技能模板" [shape=box, style=filled, fillcolor="#fff9c4"];
    "推荐技能链" [shape=box, style=filled, fillcolor="#f8bbd0"];
    "用户确认" [shape=diamond, style=filled, fillcolor="#e1bee7"];
    "执行技能链" [shape=box, style=filled, fillcolor="#81c784"];
    "保存工件" [shape=box, style=filled, fillcolor="#81c784"];

    "用户请求" -> "分析任务类型";
    "分析任务类型" -> "匹配技能模板";
    "匹配技能模板" -> "推荐技能链";
    "推荐技能链" -> "用户确认";
    "用户确认" -> "执行技能链" [label="确认"];
    "执行技能链" -> "保存工件";
}

Detailed Steps

Step 1: Analyze Task Type

Identify the task type:

🎯 Clear goal → goal-oriented
💡 Requires innovation → first-principles
🏗️ Architecture design → ddd-strategic-design
🔧 Implementation details → ddd-tactical-design
📦 Complex system → mvp-first
🔄 Continuous improvement → pdca-cycle
📊 Strategic analysis → swot-analysis

Step 2: Match Skill Templates

Based on the task type, match preset skill templates:

Template 1: New System Design

goal-oriented → first-principles → ddd-strategic-design → ddd-tactical-design → mvp-first → pdca-cycle

Template 2: Performance Optimization

goal-oriented → first-principles → pdca-cycle

Template 3: Technology Selection

goal-oriented → first-principles → swot-analysis

Template 4: Architecture Refactoring

goal-oriented → first-principles → ddd-strategic-design → pdca-cycle

Step 3: Recommend Skill Chain

Based on the user request and task type, recommend the skill execution order.

Step 4: User Confirmation

Use AskUserQuestion to let the user confirm the recommended skill chain.

Step 5: Execute Skill Chain

Execute skills inline in order (skip duplicate prerequisite protocols).

Step 6: Save Artifacts

Save the orchestration results and execution records.

Skill Template Library

Template 1: New System Design (Complete Process)

Applicable: The user says "design an XX system", "build an XX from scratch"

1. /goal-oriented
   - 明确系统目标和成功标准

2. /first-principles
   - 从本质思考系统设计

3. /ddd-strategic-design
   - 识别限界上下文
   - 设计上下文关系

4. /ddd-tactical-design
   - 设计聚合、实体、值对象
   - 实现领域模型

5. /mvp-first
   - 筛选 MVP 功能
   - 优先级规划

6. /pdca-cycle
   - 迭代实施

Template 2: Performance Optimization

Applicable: The user says "optimize XX performance", "improve XX speed"

1. /goal-oriented
   - 明确优化目标（响应时间 < 100ms）

2. /first-principles
   - 找到性能瓶颈的本质原因

3. /pdca-cycle
   - Plan: 制定优化方案
   - Do: 实施优化
   - Check: 验证性能提升
   - Act: 标准化成功做法

Template 3: Technology Selection

Applicable: The user says "choose XX technology", "XX vs YY, which is better"

1. /goal-oriented
   - 明确选型目标和约束条件

2. /first-principles
   - 从本质分析需求

3. /swot-analysis
   - 分析各方案的优劣势
   - 生成决策建议

Template 4: Architecture Refactoring

Applicable: The user says "refactor XX architecture", "optimize system structure"

1. /goal-oriented
   - 明确重构目标

2. /first-principles
   - 找到当前架构的根本问题

3. /ddd-strategic-design
   - 重新设计限界上下文
   - 优化上下文关系

4. /pdca-cycle
   - 逐步迁移
   - 持续验证

Template 5: Feature Development (Fast Path)

Applicable: The user says "implement XX feature", "add XX"

1. /goal-oriented
   - 明确功能目标

2. /mvp-first
   - 确定核心功能
   - 削减非必要特性

3. /pdca-cycle
   - 迭代开发

Template 6: Problem Diagnosis

Applicable: The user says "XX has a problem", "troubleshoot XX bug"

1. /goal-oriented
   - 明确问题解决目标

2. /first-principles
   - 找到问题的根本原因

3. /pdca-cycle
   - 实施修复
   - 验证解决方案

Skill Collaboration Rules

Inline Execution Protocol

When the orchestrator executes a skill chain:

Read The tool reads the skill file
```
skills/{skill-name}/SKILL.md
```
Skip the following parts(to avoid duplicate initialization):
- Pre-protocol (environment detection, pre-skill check)
- Overview
- When to Use
Execute the core process：
- The Process
- Thinking Framework
- Necessary interactions and decisions
Artifact output：
- Save artifact to memory/artifacts/{skill-name}/
- Continue to the next skill

Artifact Passing Mechanism

After each skill finishes executing, the artifact will be read by the next skill:

# 当前技能读取前置技能的工件
PREV_ARTIFACT="memory/artifacts/{previous-skill}/latest.json"

if [ -f "$PREV_ARTIFACT" ]; then
  # 提取上下文信息
  # 在当前技能中使用
fi

Examples

Case 1: Designing a New System from Scratch

用户："设计一个电商平台"

编排器分析：
- 任务类型：新系统设计
- 推荐模板：模板 1（完整流程）

推荐技能链：
1. /goal-oriented - 明确系统目标
2. /first-principles - 从本质思考电商
3. /ddd-strategic-design - 战略设计
4. /ddd-tactical-design - 战术设计
5. /mvp-first - MVP 规划

是否开始执行？
- A) 按推荐顺序执行
- B) 我自己选择技能
- C) 只运行第一个技能

Case 2: Performance Optimization

用户："数据库查询太慢，帮我优化"

编排器分析：
- 任务类型：性能优化
- 推荐模板：模板 2（性能优化）

推荐技能链：
1. /goal-oriented - 明确优化目标（响应时间 < 100ms）
2. /first-principles - 找到慢查询的本质原因
3. /pdca-cycle - PDCA 循环优化

是否开始执行？
- A) 按推荐顺序执行
- B) 我自己选择技能
- C) 只运行第一个技能

Case 3: Technology Selection

用户："选择微服务框架，Spring Cloud 还是 Dubbo？"

编排器分析：
- 任务类型：技术选型
- 推荐模板：模板 3（技术选型）

推荐技能链：
1. /goal-oriented - 明确选型目标
2. /first-principles - 从本质分析需求
3. /swot-analysis - 分析各方案优劣势

是否开始执行？
- A) 按推荐顺序执行
- B) 我自己选择技能
- C) 只运行第一个技能

Post-protocol

Artifact output

Save the orchestration result to the artifact file:

TIMESTAMP=$(date +%Y%m%d-%H%M%S)
ARTIFACT_FILE="memory/artifacts/pilot/result-$TIMESTAMP.json"

cat > "$ARTIFACT_FILE" <<EOF
{
  "skill": "pilot",
  "version": "1.0.0",
  "timestamp": "$(date -u +%Y-%m-%dT%H:%M:%SZ)",
  "project": "$PROJECT_ROOT",
  "branch": "$BRANCH",
  "commit": "$COMMIT",
  "input": {
    "user_request": "用户的原始请求"
  },
  "output": {
    "task_type": "识别的任务类型",
    "recommended_skills": [
      "skill-1",
      "skill-2"
    ],
    "execution_result": "success"
  }
}
EOF

echo "ARTIFACT SAVED: $ARTIFACT_FILE"
ln -sf "$ARTIFACT_FILE" memory/artifacts/pilot/latest.json

Target File Update

If the target file exists, record the orchestration completion.

Recommended Follow-up Skills

After the orchestrator completes, the skill chain has finished executing, and follow-up skills usually do not need to be recommended.

Common Pitfalls

Pitfall 1: Over-orchestration

Manifestation: Recommending complex skill chains for simple tasks

Correct approach：

Simple task → short skill chain (1-2 skills)
Complex task → complete skill chain (5-6 skills)

Pitfall 2: Ignoring User Intent

Manifestation: Mechanically recommending templates without considering the user's specific needs

Correct approach：

Analyze the details of the user's request
Adjust the recommended skill chain according to the specific needs

Pitfall 3: Forced Execution

Manifestation: Directly executing the skill chain without giving the user a chance to choose

Correct approach：

Use AskUserQuestion to let the user confirm
Provide custom options

References

gstack skill orchestration architecture
superpowers skill collaboration mode
methodolog-skills skill dependency graph

About this skill

Prerequisite Protocol

Environment Detection

Artifact Directory Initialization

Experience Retrieval (Recommended)

Error Prevention (Automatic)

Overview

When to Use

The Process

Detailed Steps

Skill Template Library

Template 1: New System Design (Complete Process)

Template 2: Performance Optimization

Template 3: Technology Selection

Template 4: Architecture Refactoring

Template 5: Feature Development (Fast Path)

Template 6: Problem Diagnosis

Skill Collaboration Rules

Inline Execution Protocol

Artifact Passing Mechanism

Examples

Case 1: Designing a New System from Scratch

Case 2: Performance Optimization

Case 3: Technology Selection

Post-protocol

Artifact output

Target File Update

Recommended Follow-up Skills

Common Pitfalls

Pitfall 1: Over-orchestration

Pitfall 2: Ignoring User Intent

Pitfall 3: Forced Execution

References