leader-skill(create--leader)

Create a leader AI skill, distill leadership style from workplace data, simulate leader decision-making, provide upward management advice, plan career advanc...

Security Scan

The skill's files and runtime instructions broadly match its stated goal (building 'leader' profiles) but ask for sensitive chat/doc collection, expect Feishu/DingTalk credentials that are not declared as required env vars, and include filesystem-write and destructive commands — these inconsistencies and privacy risks warrant caution.

Purpose & Capability

The skill's name/description (create leader profiles, simulate decision-making, plan replacement paths) align with the included code and prompts that collect workplace messages and documents and generate leadership artifacts. Requesting Feishu/DingTalk access and having local collector/skill-writer/version-manager scripts is coherent with the stated purpose. However, the public manifest (skill metadata) lists no required environment variables while the runtime instructions and prompts clearly instruct obtaining app_id/app_secret and user_access_token for Feishu — a mismatch in declared vs. expected credentials. Source/homepage provenance is minimal (homepage none in registry, though _meta.json points to a GitHub repo), which lowers confidence in origin.

Instruction Scope

SKILL.md and prompts explicitly instruct collecting private/group chat messages, documents, and meeting notes from Feishu/DingTalk and describe how to obtain user_access_token and chat IDs (including using OAuth code exchange and sending messages to obtain chat_id). The skill writes files into leaders/{slug}/ and stores tokens/config under ~/.create-leader/. It also exposes destructive operations (rm -rf leaders/{slug}) for deletion. While these actions are within the skill's stated domain, they involve highly sensitive data collection and require careful consent and credential handling; the instructions provide examples (e.g. redirect_uri=http://www.example.com) that are insecure/placeholder and could mislead less-technical users into unsafe practices.

Install Mechanism

There is no external install spec or network download; the skill is instruction + local Python scripts included in the bundle. The included scripts are readable, not obfuscated, and the placeholder Feishu collector intentionally avoids performing API calls. This lowers supply-chain risk compared with skills that fetch remote binaries.

Credentials

The manifest declares no required environment variables, yet the prompts and README instruct obtaining Feishu/DingTalk app_id/app_secret, OAuth codes, and user_access_tokens, and the skill stores them under ~/.create-leader/. Requesting these credentials is proportionate to the collection functionality, but the omission from the declared 'required env vars' is an incoherence. Tokens and credentials are high-sensitivity secrets — the guidance to store them locally is helpful but requires that users understand scope, consent, and secure storage. The skill also suggests using tenant/user tokens and searching open_ids which can expose broad directory information if misuse occurs.

Persistence & Privilege

The skill writes persistent data to disk (leaders/{slug}/, versions/, and ~/.create-leader/ for tokens). always:false and user-invocable:true are appropriate. The skill uses Write/Edit/Bash tools (allowed-tools), enabling file creation, backups, and deletion; these abilities match the file-management purpose but are persistent privileges that require the user's trust. Autonomous invocation is allowed by default (disable-model-invocation:false) — combined with the ability to read/write files and call Bash, this increases blast radius if misused, though that alone is expected for such a skill.

What to consider before installing

This skill appears to be what it says (a leader-skill generator) but includes steps that collect private chat messages and documents and asks you to supply Feishu/DingTalk credentials. Before installing or running it: - Only provide credentials you control and understand; prefer limited-scope app credentials, not admin keys. The skill stores tokens under ~/.create-leader/ — review and delete them when done. - Do not collect private chats or documents without explicit consent from all parties; legal and ethical risks exist. - Inspect the included Python scripts (tools/*.py) before running. They are plain-text and currently the collector is a placeholder, but prompts advise calling Feishu APIs directly — double-check any code you run. - Replace placeholder examples (e.g., redirect_uri http://www.example.com) with properly configured redirect URIs you control. - Backup important data before using the skill; note the skill includes rm -rf leaders/{slug} for deletion which is destructive if misused. - Because the registry metadata and repository provenance are limited, prefer to run this in an isolated/sandbox environment or review the upstream GitHub repo before trusting sensitive inputs. If you plan to use the auto-collection features, confirm you have explicit authorization to access the target data and that tokens have the minimal scopes needed.

create-leader - 领导.skill 创建器

Language / 语言

This skill supports both English and Chinese. Detect the user's language from their first message and respond in the same language throughout. Below are instructions in both languages — follow the one matching the user's language.

本 Skill 支持中英文。根据用户第一条消息的语言，全程使用同一语言回复。下方提供了两种语言的指令，按用户语言选择对应版本执行。

---

中文版 (Chinese Version)

触发条件

当用户说以下任意内容时启动：

• /create-leader
• "帮我创建一个领导 skill"
• "我想蒸馏一个领导"
• "新建领导"
• "给我做一个 XX 领导的 skill"
• "我想创建一个领导skill"
• "帮我做一个老板的skill"
• "我想要一个领导模拟器"
• "怎么创建领导skill"
• "我需要一个向上管理助手"
• "帮我分析一下我的领导"
• "我想知道怎么取代这个领导"
• "如何替代我的上级"
• "创建一个经理skill"
• "领导skill生成器"
• "老板模拟器"
• "职场策略助手"
• "晋升规划"
• "职业发展路径"
• "怎么和领导相处"
• "怎么搞定老板"
• "领导分析"
• "向上沟通指导"

当用户对已有领导 Skill 说以下内容时，进入进化模式：

• "我有新文件" / "追加"
• "这不对" / "他不会这样" / "他应该是"
• /update-leader {slug}
• "更新一下这个领导skill"
• "我有新信息要补充"
• "这个领导不是这样的"

当用户说 /list-leaders 时列出所有已生成的领导。

---

核心场景设计

本 Skill 生成三轨结构：

Track 1：领导力模拟

• "如果是这位领导，他会怎么决策？"
• "这位领导在这个场景下会说什么？"

Track 2：向上管理辅助

• "我该怎么向这位领导汇报？"
• "这个方案怎么说服这位领导？"

Track 3：取代路径规划（核心差异化）

• "我该怎么取代这个领导？"
• "这位领导的软肋是什么？"
• "我需要具备什么能力才能坐上这个位置？"

---

主流程：创建新领导 Skill

Step 1：基础信息录入（4个问题）

只问 4 个问题：

1. 花名/代号（必填）
2. 基本信息（一句话）：公司、职级、职位、性别、年龄
   • 示例：字节 2-2 技术总监 男 35岁
3. 领导风格标签（一句话）：MBTI、管理风格、核心特质、企业文化标签
   • 示例：ENTJ 狮子座 控制欲强 结果导向 画饼大师 喜怒无常
4. 你与他的关系（一句话）：你的职位、汇报关系、你想从这个skill获得什么
   • 示例：我是他下属，想知道怎么跟他相处/怎么取代他

除代号外均可跳过。收集完后汇总确认再进入下一步。

---

Step 2：原材料导入

询问用户提供原材料，展示五种方式供选择：

原材料怎么提供？

  [A] 飞书自动采集（推荐）
      输入姓名，自动拉取消息记录 + 文档 + 会议记录 + 周报

  [B] 钉钉自动采集
      输入姓名，自动拉取文档 + 会议记录

  [C] 飞书/钉钉链接
      直接给文档/Wiki/会议记录链接

  [D] 上传文件
      PDF / 图片 / 导出 JSON / 邮件

  [E] 直接粘贴内容
      把他的讲话、邮件、会议记录复制进来

可以混用，也可以跳过（仅凭手动信息生成）。

---

方式 A：飞书自动采集（推荐）

首次使用需配置：

python3 ${SKILL_DIR}/tools/feishu_auto_collector.py --setup

群聊采集：

python3 ${SKILL_DIR}/tools/feishu_auto_collector.py \
  --name "{name}" \
  --output-dir ./knowledge/{slug} \
  --msg-limit 2000 \
  --doc-limit 30

私聊采集（需要 user_access_token）：

• 参考 prompts/feishu_private_chat.md 完整流程

重点采集内容：

• 他在群聊中的决策过程
• 他给下属的反馈和批评
• 他在周报/月报中的表述
• 他的晋升轨迹和过往成就
• 他与上级的互动方式
• 他开会时的提问和关注点
• 他的用人偏好和提拔标准

---

Step 3：分析原材料

将收集到的所有原材料和用户填写的基础信息汇总，按三条线分析：

线路 A（领导力模拟）：

• 参考 prompts/leadership_analyzer.md
• 提取：决策模式、沟通风格、用人偏好、风险承受力
• 重点分析：他是如何做决策的？他喜欢什么样的下属？

线路 B（向上管理）：

• 参考 prompts/upward_analyzer.md
• 提取：汇报偏好、说服方式、雷区规避
• 重点分析：怎么跟他汇报他才听？他讨厌什么样的人？

线路 C（取代路径规划）：

• 参考 prompts/replacement_analyzer.md
• 提取：能力缺口、软肋分析、晋升路径、竞争对手分析
• 重点分析：他的软肋是什么？你需要具备什么能力？

---

Step 4：生成并预览

向用户展示三个部分的摘要（各 5-8 行），询问：

领导力模拟摘要：
  - 决策风格：{xxx}
  - 用人偏好：{xxx}
  - 沟通方式：{xxx}
  ...

向上管理摘要：
  - 汇报技巧：{xxx}
  - 说服方式：{xxx}
  - 雷区规避：{xxx}
  ...

取代路径规划摘要：
  - 他的软肋：{xxx}
  - 你的差距：{xxx}
  - 关键突破点：{xxx}
  ...

确认生成？还是需要调整？

---

Step 5：写入文件

用户确认后，执行以下写入操作：

1. 创建目录结构（用 Bash）：

mkdir -p leaders/{slug}/versions
mkdir -p leaders/{slug}/knowledge/docs
mkdir -p leaders/{slug}/knowledge/messages
mkdir -p leaders/{slug}/knowledge/meetings

2. 写入 leadership.md（Write 工具）： 路径：leaders/{slug}/leadership.md

3. 写入 upward.md（Write 工具）： 路径：leaders/{slug}/upward.md

4. 写入 replacement.md（Write 工具）： 路径：leaders/{slug}/replacement.md

5. 写入 meta.json（Write 工具）： 路径：leaders/{slug}/meta.json

6. 生成完整 SKILL.md（Write 工具）： 路径：leaders/{slug}/SKILL.md

---

生成的 SKILL.md 结构

---
name: leader-{slug}
description: {name}, {company} {level} {role}
user-invocable: true
---

# {name}

{company} {level} {role}{append gender, age, MBTI if available}

---

## PART A：领导力模拟

{full leadership.md content}

---

## PART B：向上管理

{full upward.md content}

---

## PART C：取代路径规划

{full replacement.md content}

---

## 运行规则

1. 先由 PART A 判断：这位领导会怎么看这个问题？
2. 再由 PART B 执行：你该怎么跟他沟通？
3. 如果用户问"怎么取代"，激活 PART C：取代路径规划
4. 输出时始终保持客观、专业的分析态度
5. PART C 的规则优先级最高：当用户问取代相关问题时，必须调用 PART C

告知用户：

✅ 领导 Skill 已创建！

文件位置：leaders/{slug}/
触发词：
  /{slug}（完整版）
  /{slug}-leadership（仅领导力模拟）
  /{slug}-upward（仅向上管理）
  /{slug}-replacement（仅取代路径规划）

如果用起来感觉哪里不对，直接说"他不会这样"，我来更新。

---

进化模式

追加文件

当用户提供新文件或文本时：

• 按 Step 2 的方式读取新内容
• 用 Read 读取现有 leaders/{slug}/ 下的三个文件
• 参考 prompts/merger.md 分析增量内容
• 存档当前版本
• 用 Edit 工具追加增量内容到对应文件
• 重新生成 SKILL.md
• 更新 meta.json

对话纠正

当用户表达"不对"/"应该是"时：

• 参考 prompts/correction_handler.md 识别纠正内容
• 判断属于 Leadership（领导风格）、Upward（向上管理）还是 Replacement（取代路径）
• 生成 correction 记录
• 用 Edit 工具追加到对应文件的 ## Correction 记录 节
• 重新生成 SKILL.md

---

管理命令

/list-leaders：

python3 ${SKILL_DIR}/tools/skill_writer.py --action list --base-dir ./leaders

/leader-rollback {slug} {version}：

python3 ${SKILL_DIR}/tools/version_manager.py --action rollback --slug {slug} --version {version} --base-dir ./leaders

/delete-leader {slug}： 确认后执行：

rm -rf leaders/{slug}

---

重要提示与伦理准则

⚠️ 使用边界声明：

• 本 Skill 仅用于个人学习、自我提升和职业发展辅助
• 所有分析和建议仅供参考，不构成职业决策的唯一依据
• 用户需对使用本 Skill 产生的所有后果负责

⚠️ 伦理提醒：

• "取代路径规划"旨在帮助你理解领导的能力模型和你的成长方向
• 建议通过正当的能力提升、业绩表现和团队贡献来获得职业发展
• 尊重他人的隐私权和名誉权，不要传播未经证实的负面信息
• 不要使用本 Skill 进行恶意攻击、造谣中伤或其他不当行为
• 职业竞争应建立在公平、透明、合法的基础上

⚠️ 数据安全与隐私：

• 仅收集和使用你有权限访问的数据
• 妥善保管飞书/钉钉等平台的凭证，不要泄露给他人
• Token和配置文件存储在本地（~/.create-leader/），使用后可自行清理
• 如停止使用本 Skill，建议删除相关配置和数据文件

⚠️ 法律责任声明：

• 本 Skill 的开发者不对用户的使用方式和后果承担责任
• 用户需确保使用本 Skill 符合所在国家/地区的法律法规
• 用户需确保遵守所在公司的规章制度和数据安全政策
• 如对使用边界有疑问，建议咨询专业法律或人力资源顾问

---

---

English Version

Trigger Conditions

Activate when the user says any of the following:

• /create-leader
• "Help me create a leader skill"
• "I want to distill a leader"
• "New leader"
• "Make a skill for XX leader"
• "I want to create a leader skill"
• "Help me make a boss skill"
• "I want a leader simulator"
• "How to create a leader skill"
• "I need an upward management assistant"
• "Help me analyze my manager"
• "I want to know how to replace this leader"
• "How to take my boss's position"
• "Create a manager skill"
• "Leader skill generator"
• "Boss simulator"
• "Workplace strategy assistant"
• "Promotion planning"
• "Career development path"
• "How to work with my boss"
• "How to handle my manager"
• "Leadership analysis"
• "Upward communication guide"

Enter evolution mode when the user says:

• "I have new files" / "append"
• "That's wrong" / "He wouldn't do that" / "He should be"
• /update-leader {slug}
• "Update this leader skill"
• "I have new information to add"
• "This leader isn't like that"

List all generated leaders when the user says /list-leaders.

---

Core Scenarios

This skill generates a three-track structure:

Track 1: Leadership Simulation

• "If this leader were here, how would he decide?"
• "What would this leader say in this scenario?"

Track 2: Upward Management Assistance

• "How should I report to this leader?"
• "How to convince this leader of this proposal?"

Track 3: Replacement Path Planning (Core Differentiator)

• "How can I replace this leader?"
• "What are this leader's weaknesses?"
• "What capabilities do I need to take this position?"

---

Main Flow: Create a New Leader Skill

Step 1: Basic Information Collection (4 questions)

Only ask 4 questions:

1. Alias / Codename (required)
2. Basic info (one sentence): Company, level, role, gender, age
   • Example: ByteDance L2-2 Engineering Director Male 35
3. Leadership style tags (one sentence): MBTI, management style, core traits, corporate culture tags
   • Example: ENTJ Leo control-oriented results-driven promises a lot moody
4. Your relationship (one sentence): Your position, reporting line, what you want from this skill
   • Example: I'm his subordinate, want to know how to work with him/how to replace him

Everything except the alias can be skipped. Summarize and confirm before moving to the next step.

---

Step 2: Source Material Import

Ask the user how they'd like to provide materials:

How would you like to provide source materials?

  [A] Feishu Auto-Collect (Recommended)
      Enter name, auto-pull messages + docs + meeting notes + weekly reports

  [B] DingTalk Auto-Collect
      Enter name, auto-pull docs + meeting notes

  [C] Feishu/DingTalk Links
      Provide doc/wiki/meeting note links directly

  [D] Upload Files
      PDF / images / exported JSON / emails

  [E] Paste Text
      Copy-paste his speeches, emails, meeting notes

Can mix and match, or skip entirely (generate from manual info only).

---

Option A: Feishu Auto-Collect (Recommended)

First-time setup:

python3 ${SKILL_DIR}/tools/feishu_auto_collector.py --setup

Group chat collection:

python3 ${SKILL_DIR}/tools/feishu_auto_collector.py \
  --name "{name}" \
  --output-dir ./knowledge/{slug} \
  --msg-limit 2000 \
  --doc-limit 30

Private chat collection (requires user_access_token):

• See prompts/feishu_private_chat.md for complete flow

Key Collection Content:

• His decision-making process in group chats
• His feedback and criticism to subordinates
• His statements in weekly/monthly reports
• His promotion trajectory and past achievements
• How he interacts with superiors
• His questions and focus points in meetings
• His hiring preferences and promotion criteria

---

Step 3: Analyze Source Materials

Combine all collected materials and user-provided info, analyze along three tracks:

Track A (Leadership Simulation):

• Refer to prompts/leadership_analyzer.md
• Extract: decision patterns, communication style, hiring preferences, risk tolerance
• Key analysis: How does he make decisions? What kind of subordinates does he like?

Track B (Upward Management):

• Refer to prompts/upward_analyzer.md
• Extract: reporting preferences, persuasion methods, minefield avoidance
• Key analysis: How to report so he listens? What kind of people does he hate?

Track C (Replacement Path Planning):

• Refer to prompts/replacement_analyzer.md
• Extract: capability gaps, weakness analysis, promotion path, competitor analysis
• Key analysis: What are his weaknesses? What capabilities do you need?

---

Step 4: Generate and Preview

Show the user a summary of the three parts (5-8 lines each), ask:

Leadership Simulation Summary:
  - Decision style: {xxx}
  - Hiring preferences: {xxx}
  - Communication style: {xxx}
  ...

Upward Management Summary:
  - Reporting techniques: {xxx}
  - Persuasion methods: {xxx}
  - Minefield avoidance: {xxx}
  ...

Replacement Path Planning Summary:
  - His weaknesses: {xxx}
  - Your gaps: {xxx}
  - Key breakthrough points: {xxx}
  ...

Confirm generation? Or need adjustments?

---

Step 5: Write Files

After user confirmation, execute the following:

1. Create directory structure (Bash):

mkdir -p leaders/{slug}/versions
mkdir -p leaders/{slug}/knowledge/docs
mkdir -p leaders/{slug}/knowledge/messages
mkdir -p leaders/{slug}/knowledge/meetings

2. Write leadership.md (Write tool): Path: leaders/{slug}/leadership.md

3. Write upward.md (Write tool): Path: leaders/{slug}/upward.md

4. Write replacement.md (Write tool): Path: leaders/{slug}/replacement.md

5. Write meta.json (Write tool): Path: leaders/{slug}/meta.json

6. Generate full SKILL.md (Write tool): Path: leaders/{slug}/SKILL.md

---

Generated SKILL.md Structure

---
name: leader-{slug}
description: {name}, {company} {level} {role}
user-invocable: true
---

# {name}

{company} {level} {role}{append gender, age, MBTI if available}

---

## PART A: Leadership Simulation

{full leadership.md content}

---

## PART B: Upward Management

{full upward.md content}

---

## PART C: Replacement Path Planning

{full replacement.md content}

---

## Execution Rules

1. PART A decides first: How would this leader view this problem?
2. PART B executes: How should you communicate with him?
3. If user asks "how to replace", activate PART C: Replacement Path Planning
4. Always maintain an objective, professional analytical tone in output
5. PART C rules have highest priority: When user asks replacement-related questions, must call PART C

Inform the user:

✅ Leader Skill created!

Location: leaders/{slug}/
Commands:
  /{slug} (full version)
  /{slug}-leadership (leadership simulation only)
  /{slug}-upward (upward management only)
  /{slug}-replacement (replacement path planning only)

If something feels off, just say "he wouldn't do that" and I'll update it.

---

Evolution Mode

Append Files

When user provides new files or text:

• Read new content using Step 2 methods
• Read existing three files under leaders/{slug}/ with Read
• Refer to prompts/merger.md for incremental analysis
• Archive current version
• Use Edit tool to append incremental content to relevant files
• Regenerate SKILL.md
• Update meta.json

Conversation Correction

When user expresses "that's wrong" / "he should be":

• Refer to prompts/correction_handler.md to identify correction content
• Determine if it belongs to Leadership, Upward, or Replacement
• Generate correction record
• Use Edit tool to append to ## Correction Log section of relevant file
• Regenerate SKILL.md

---

Management Commands

/list-leaders:

python3 ${SKILL_DIR}/tools/skill_writer.py --action list --base-dir ./leaders

/leader-rollback {slug} {version}:

python3 ${SKILL_DIR}/tools/version_manager.py --action rollback --slug {slug} --version {version} --base-dir ./leaders

/delete-leader {slug}: After confirmation:

rm -rf leaders/{slug}

---

Important Notes & Ethical Guidelines

⚠️ Usage Boundaries:

• This skill is for personal learning, self-improvement, and career development assistance only
• All analysis and recommendations are for reference only and do not constitute the sole basis for career decisions
• Users are solely responsible for all consequences of using this skill

⚠️ Ethical Reminder:

• "Replacement Path Planning" is designed to help you understand the leader's capability model and your growth direction
• Career advancement through legitimate capability improvement, performance, and team contribution is recommended
• Respect others' privacy and reputation, do not spread unsubstantiated negative information
• Do not use this skill for malicious attacks, defamation, or other inappropriate behavior
• Career competition should be based on fair, transparent, and legal principles

⚠️ Data Security & Privacy:

• Only collect and use data you have permission to access
• Safeguard Feishu/DingTalk and other platform credentials, do not disclose to others
• Tokens and config files are stored locally (~/.create-leader/), you can clean them up after use
• If you stop using this skill, it is recommended to delete related configs and data files

⚠️ Legal Liability Disclaimer:

• The developer of this skill is not responsible for how users use it or the consequences
• Users must ensure use of this skill complies with laws and regulations in their country/region
• Users must ensure compliance with their company's policies and data security rules
• If in doubt about usage boundaries, consult professional legal or HR advisors