Feishu Requirement Manager

飞书需求管理工具（去中心化+可扩展）。支持需求创建、任务拆分、代理执行、进度跟踪。 **核心字段**： - 需求：标题、状态、优先级、进度、截止日期 - 任务：标题、状态、执行人、关联需求 **支持自定义扩展**，用户可添加额外字段。

Security Scan

The skill's instructions and requirements are internally consistent with being a Feishu (Bitable) requirement manager, but there are a few ambiguities around authentication and persistent token storage you should check before installing.

Purpose & Capability

The SKILL.md describes creating projects/requirements/tasks and mapping fields to Feishu Bitable tables, and the examples show creating app/table/record objects via feishu_bitable_app.* calls — this aligns with the stated purpose.

Instruction Scope

Instructions explicitly read/write a memory file (memory/shared/requirement-manager.json) to persist app_token and table IDs and direct agents to create/read/update Bitable records. Persisting an app_token is expected for continued access, but the SKILL.md does not detail how Feishu authentication is obtained or what permissions the token has — worth verifying. The skill also delegates work to other agents (PM agent, execution agents), which is consistent with a decentralized workflow but broadens runtime actions.

Install Mechanism

This is an instruction-only skill with no install spec or code files. Nothing is written to disk by an installer here beyond the agent's normal memory persistence described in the instructions.

Credentials

The registry metadata declares no required env vars or credentials, and SKILL.md relies on an app_token produced/used at runtime. That is coherent if the platform provides a Feishu connector, but the skill doesn't document how the token is obtained or what privileges it needs. Storing app_token in agent memory persists a credential — confirm token scope and lifecycle.

Persistence & Privilege

always is false and the skill does not request system-wide modification. It does instruct writing to a shared memory file to persist app_token and table IDs, which grants it persistent access to those Feishu resources; ensure that stored tokens are scoped minimally and that you are comfortable with agent memory storage of credentials.

Assessment

This skill appears to do what it claims: manage Feishu Bitable requirement and task records. Before installing, verify these points: (1) Authentication model — confirm how feishu_bitable_app connector obtains the app_token and whether you must provide Feishu credentials elsewhere; (2) Token scope and rotation — ensure the stored app_token has only the permissions needed and can be revoked/rotated; (3) Memory storage location — confirm where memory/shared/requirement-manager.json is stored, who can access it, and whether it will persist across sessions; (4) Least privilege for created resources — the tables/apps it creates should not require excessive org-wide privileges; (5) Delegation behavior — the skill delegates tasks to other agents, so review what those agents are allowed to do in your environment. If you cannot confirm the connector/auth details, treat the persisted app_token as a sensitive secret and avoid installing until clarified.

需求管理 Skill (v3.1.0)

📋 核心字段定义

需求列表（必填字段）

字段	类型	说明
需求标题	text	必填，唯一标识
状态	singleSelect	待处理/进行中/已完成/已取消
优先级	singleSelect	P0/P1/P2/P3
进度	number	0-100%
截止日期	datetime	可选
提出日期	datetime	自动填充

需求列表（可扩展字段）

字段	类型	说明
需求描述	text	详细描述
需求类型	singleSelect	产品/技术/运营/其他
所属项目	link	关联项目
关联文档	url	相关文档链接
负责人	user	负责人
备注	text	备注信息

---

任务列表（必填字段）

字段	类型	说明
任务标题	text	必填
任务状态	singleSelect	待处理/进行中/已完成/阻塞
关联需求	link	关联需求（必填）
执行人	text/select	AI助手/搜索Agent/文档Agent/用户

任务列表（可扩展字段）

字段	类型	说明
任务描述	text	详细描述
依赖任务	link	依赖的其他任务
优先级	singleSelect	P0/P1/P2/P3
截止日期	datetime	截止时间
输出	text	执行结果/交付物

---

项目列表（可选）

字段	类型	说明
项目名称	text	必填
项目状态	singleSelect	进行中/已完成/已暂停
项目描述	text	描述
负责人	user	项目负责人

---

🔄 动态字段处理

创建时的字段映射

// 用户输入 → 字段映射
用户输入 = {
  "需求": "开发一个App",
  "优先级": "P1",
  "类型": "产品"
}

// 转换为表格字段
fields = {
  "需求标题": "开发一个App",
  "优先级": "P1 - 重要",
  "需求类型": "产品"
}

自定义字段处理

如果用户添加了额外字段：

1. 尝试匹配已知字段
2. 无法匹配时，提示用户确认字段名
3. 存储到"备注"字段或创建新字段

---

⚡ 核心流程

用户请求
    ↓
主Agent：创建需求/任务
    ↓
PM代理：拆分任务
    ↓
执行Agent：并行执行
    ↓
更新表格状态
    ↓
通知用户

---

📖 使用示例

例1: 创建需求

用户：创建需求：开发AI助手

主Agent：
1. 检查表格是否存在
2. 创建需求记录
   - 需求标题：开发AI助手
   - 状态：待处理
   - 优先级：P1 - 重要
   - 提出日期：当前时间
3. 触发PM代理

例2: 添加自定义字段

用户：创建需求：项目A，预算10万，负责人张三

主Agent：
1. 核心字段：需求标题、项目A
2. 自定义字段：
   - 预算：10万 → 存到"备注"
   - 负责人：张三 → 如有"负责人"字段则使用，否则存备注

---

🔗 触发关键词

需求管理、创建需求、拆分需求、执行任务
查看进度、添加任务

---

⚠️ 字段映射规则

用户输入	表格字段
需求/标题	需求标题
任务	任务标题
紧急/重要	P0 - 紧急 / P1 - 重要
正在进行/做	进行中
完成/结束	已完成

---

版本: 3.1.0 | 2026-03-22