first-principles — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited first-principles (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
# 检测当前项目信息
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: $COMMIT"benefits-from 检查(推荐但非必须):
# 检查 goal-oriented 工件
GOAL_ARTIFACT="memory/artifacts/goal-oriented/latest.json"
if [ -f "$GOAL_ARTIFACT" ]; then
echo "FOUND: goal-oriented artifact"
# 提取目标信息(使用 Read 工具读取)
# 在分析中参考目标上下文
else
echo "INFO: No goal-oriented artifact found"
echo "Consider running /goal-oriented first for better context"
fi工件目录初始化:
# 确保工件目录存在
mkdir -p memory/artifacts/first-principles根据用户消息判断:
检查点:
意图分类:
第一性原理是一种从最基础、最根本的真理或事实出发,重新构建问题解决方案的思维方式。它要求抛开现有假设、惯例或类比,直接追问"这件事的本质是什么?""最基本的构成要素是什么?",然后基于这些基础元素推导出新的可能性。
核心区别:
典型例子: 埃隆·马斯克思考火箭制造成本时,不是接受市场价,而是从原材料成本出发,得出自己制造更便宜的结论。
适用场景:
不适用场景:
digraph first_principles {
rankdir=TB;
"识别问题" [shape=box, style=filled, fillcolor="#e0e0e0"];
"拆解到基础元素" [shape=box, style=filled, fillcolor="#c8e6c9"];
"质疑假设" [shape=box, style=filled, fillcolor="#fff9c4"];
"重建方案" [shape=box, style=filled, fillcolor="#bbdefb"];
"验证可行性" [shape=box, style=filled, fillcolor="#f8bbd0"];
"保存工件" [shape=box, style=filled, fillcolor="#81c784"];
"识别问题" -> "拆解到基础元素";
"拆解到基础元素" -> "质疑假设";
"质疑假设" -> "重建方案";
"重建方案" -> "验证可行性";
"验证可行性" -> "识别问题" [label="不可行", style=dashed];
"验证可行性" -> "保存工件" [label="可行"];
}步骤 1: 识别问题
步骤 2: 拆解到基础元素
步骤 3: 质疑假设
步骤 4: 重建方案
步骤 5: 验证可行性
步骤 6: 保存工件
使用以下表格系统化地进行第一性原理思考:
| 层级 | 问题 | 提示 | 示例(火箭成本) |
|---|---|---|---|
| 表象 | 现在的问题是什么? | 描述症状和现象 | 火箭太贵,$65M/个 |
| 假设 | 我们认为的"必须如此"是什么? | 列出所有假设 | 必须买现成的、供应商定价合理 |
| 本质 | 最基础的构成要素是什么? | 物理定律、原材料、基本原理 | 铝、钛、碳纤维,材料成本$800K |
| 重建 | 如何从本质重新构建? | 忽略现有方案,从头设计 | 自己制造,成本降至$8M |
思考提示:
问题: 查询太慢,平均耗时 10 秒
表象: 查询慢,需要优化 SQL
假设:
本质:
重建:
问题: 需要选择微服务框架
表象: Spring Cloud、Dubbo、gRPC 哪个更好?
假设:
本质:
重建:
保存第一性原理分析结果到工件文件:
# 生成工件文件名
TIMESTAMP=$(date +%Y%m%d-%H%M%S)
ARTIFACT_FILE="memory/artifacts/first-principles/result-$TIMESTAMP.json"
# 写入工件
cat > "$ARTIFACT_FILE" <<EOF
{
"skill": "first-principles",
"version": "2.0.0",
"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": {
"problem": "识别的问题",
"assumptions": [
"假设1",
"假设2"
],
"fundamentals": [
"基础元素1",
"基础元素2"
],
"solution": "重建的解决方案",
"validation": "可行性验证结果"
},
"next_skills": [
"ddd-strategic-design",
"mvp-first",
"pdca-cycle"
]
}
EOF
echo "ARTIFACT SAVED: $ARTIFACT_FILE"
# 创建 latest.json 符号链接
ln -sf "$ARTIFACT_FILE" memory/artifacts/first-principles/latest.json如果存在目标文件,记录分析完成:
# 检查是否有 pending 目标
GOAL_FILE=$(ls -t memory/goals/*.md 2>/dev/null | head -1)
if [ -n "$GOAL_FILE" ]; then
GOAL_STATUS=$(grep "状态:" "$GOAL_FILE" | awk '{print $2}')
if [ "$GOAL_STATUS" = "pending" ]; then
echo "GOAL STATUS: $GOAL_STATUS"
echo "Adding milestone: 第一性原理分析完成"
# 使用 Edit 工具添加里程碑
# 例如:"第一性原理分析完成 - {时间}"
fi
fi根据分析结果,推荐后续技能:
推荐格式:
## 后续建议
基于第一性原理分析结果,建议继续执行:
**推荐技能链**:
1. /ddd-strategic-design - 如果涉及系统架构设计
2. /mvp-first - 如果是新系统或新功能
3. /pdca-cycle - 进入迭代执行阶段
**根据解决方案类型选择**:
- **架构设计类** → /ddd-strategic-design → /ddd-tactical-design
- **新系统开发** → /mvp-first → /pdca-cycle
- **性能优化类** → /pdca-cycle
- **技术选型类** → 直接进入实施
是否继续执行?
- A) 执行推荐的技能链
- B) 只执行第一个技能
- C) 不继续,结束当前任务~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.