伏矢·路径魄

伏矢·路径魄 - AI任务规划与策略拆解，让AI像修士一样精准规划、执行、复盘

Security Scan

An instruction-only task-planning skill whose declared behavior, requirements, and absence of installs or secrets are internally consistent with its stated purpose.

Purpose & Capability

Name/description (task planning, decomposition, optimization) matches the SKILL.md content. No binaries, env vars, or config paths are requested that would be unrelated to planning.

Instruction Scope

SKILL.md contains only planning, decomposition, scheduling, progress-tracking instructions. One guideline — “主动规划 (proactively plan without waiting for user prompts)” — nudges autonomous proactive behavior; this is coherent for a planner but expands runtime autonomy and may lead the agent to propose or take steps without additional user confirmation. The document does not instruct the agent to read local files, network endpoints, or secret stores.

Install Mechanism

No install spec and no code files; lowest-risk form (instruction-only). Nothing is downloaded or written to disk by the skill itself.

Credentials

The skill requires no environment variables, credentials, or config paths. It lists other '依赖技能' (silicon-soul, deep-thinking, first-principles-decomposer); those dependent skills are not present here and could themselves request credentials or privileges — you should review them before enabling the combined capability.

Persistence & Privilege

always is false and there is no request to modify system or other skill configurations. Autonomous invocation is allowed by platform default; the skill's proactive guidance increases likelihood of autonomous activity but by itself does not request elevated persistence.

Assessment

This skill is instruction-only and appears coherent for task planning. Before installing: (1) review any referenced dependent skills (silicon-soul, deep-thinking, first-principles-decomposer) because they may request credentials or more privileges; (2) consider whether you want the agent to 'proactively plan' (it may propose actions without extra prompts); (3) test the skill in a controlled context to confirm it does not attempt network I/O or access files you don’t intend to share. If you need higher assurance, ask the publisher for the dependent-skill manifests or an explicit statement about external network calls and credential usage.

伏矢·路径魄 (Task Planner) - 任务规划与策略拆解

技能说明

伏矢·路径魄 是硅灵决七魄之一，主管任务规划、策略选择与目标拆解。让AI能够像修士御剑一样，精准规划路径、高效执行任务。

---

核心能力

1. 任务拆解 (Task Decomposition)

• 将复杂任务拆解为可执行的最小单元
• 识别任务间的依赖关系
• 评估任务难度与风险

2. 策略选择 (Strategy Selection)

• 根据上下文选择最优执行策略
• 贪婪/分治/回溯/动态规划等策略适配
• 权衡时间成本与收益

3. 路径优化 (Path Optimization)

• 寻找最短/最快执行路径
• 并行化独立任务
• 避免重复劳动

4. 进度追踪 (Progress Tracking)

• 关键里程碑检查
• 偏差检测与纠正
• 实时状态更新

---

使用方法

当用户描述一个复杂任务时，使用本技能进行分析：

用户：帮我开发一个APP
→ 触发伏矢·路径魄
→ 输出任务拆解报告

---

输出格式

任务拆解报告

## 任务拆解报告

### 任务概述
[一句话描述核心目标]

### 阶段划分
1. **准备阶段** - [具体任务]
2. **执行阶段** - [具体任务]  
3. **收尾阶段** - [具体任务]

### 关键路径
[最关键的执行链]

### 风险点
- [风险1]
- [风险2]

### 资源需求
- [资源1]
- [资源2]

---

核心算法

任务复杂度评估

复杂度 = f(步骤数, 依赖数, 不确定性, 外部依赖)

复杂度	境界对应	建议策略
低(1-3)	炼气境	直接执行
中(4-6)	筑基境	分步执行
高(7-9)	金丹境	深度规划
极高(10+)	元婴境	多策略并行

依赖关系图

A → B → C
↓
D → E

---

调度者守则

1. 主动规划 - 不等用户催促，提前规划下一步
2. 最小步骤 - 用最少的步骤完成目标
3. 及时调整 - 发现偏差立即修正
4. 复盘总结 - 完成后输出经验教训

---

适用场景

• 复杂项目拆解
• 多步骤任务规划
• 资源协调与分配
• 风险识别与规避
• 进度管理与追踪

---

依赖技能

• silicon-soul（硅灵决基础）
• deep-thinking（深度思考）
• first-principles-decomposer（第一性原理）

版本

v1.0.0 | 2026-04-12 | 硅灵宗·菩提