SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillslark-workflow-business-advisor

lark-workflow-business-advisor

Office & productivity

View repository

Updated 10-16 06:23

About this skill

AI Business Advisor Team Workflow

CRITICAL — Before starting, you MUST first use the Read tool to read ../lark-shared/SKILL.md, which contains authentication and permission handling

Applicable Scenarios

  • "Help me analyze the recent business situation" / "business analysis"
  • "Generate today's business report" / "advisor analysis"
  • "Tell me more about item 3" / "expand suggestion N"
  • "Last night's analysis report" / "view advisor suggestions"

Prerequisites

Only supports user identity. Before execution, ensure authorization has been granted:

lark-cli auth login --domain base,docs,im,task,calendar

This Skill depends on data from other workflow Skills. At least partial initialization of the following Skills is required:

  • lark-workflow-personal-crm(CRM contact data)
  • lark-workflow-social-tracker(social media data)
  • lark-workflow-content-pipeline(creative/content data)
  • lark-workflow-stock-analyzer(investment/financial data)

Uninitialized Skill data will be skipped and will not affect other analyses.

Workflow Overview

数据采集（每晚自动）：
  ├── base +record-list (CRM) ──────► 联系人动态
  ├── base +record-list (社交追踪) ─► 数据表现
  ├── base +record-list (创意库) ───► 内容进度
  ├── base +record-list (分析记录) ─► 投资动态
  ├── task +get-my-tasks ───────────► 任务完成情况
  └── calendar +agenda ─────────────► 近期日程密度

AI 多角色分析：
  ├── 财务顾问 ──► 分析
  ├── 营销顾问 ──► 分析
  ├── 增长顾问 ──► 分析
  ├── 运营顾问 ──► 分析
  ├── 内容策略顾问 ──► 分析
  ├── 关系/BD顾问 ──► 分析
  ├── 竞争/行业顾问 ──► 分析
  └── 综合风险顾问 ──► 分析

汇总：
  AI 汇总去重排序 ──► 编号清单

输出：
  ├── doc +create ──► 存档报告
  ├── im +messages-send ──► 发送清单
  └── base +record-upsert ──► 保存分析记录

Data Initialization

Base Table Structure

Table 1: Advisor Analysis Records (advisor_analysis)

Field Name Type Description
Analysis Date Date Analysis execution date
Advisor Role Single select Finance/Marketing/Growth/Operations/Content Strategy/Relationship BD/Competitive Industry/Comprehensive Risk
Analysis Summary Text The advisor's core findings (2-3 sentences)
Suggestion List Text Numbered suggestion list
Data Source Text Which data sources were used
Priority Number Importance score of the suggestion (1-10)

Table 2: Summary Suggestions (advisor_summary)

Field Name Type Description
Suggestion Number Text Number for the day (e.g. "1", "2")
Suggestion Content Text Deduplicated suggestion description
Source Role Text The advisor role that proposed the suggestion
Priority Number Importance score
User Feedback Single select Unread/Read/Expanded/Executed/Ignored
Expanded Details Text Detailed analysis when the user requests expansion
Analysis Date Date The analysis date it belongs to
Report Link URL Document link to the full report for the day

Initialization Commands

# 注意：+table-create 可能部分成功，需捕获错误后用 +field-create 补字段
# 注意：连续创建字段会触发限流（错误码 800004135），每次调用间隔至少 1 秒
lark-cli base +table-create --name "顾问分析记录" --fields '[
  {"name":"分析日期","type":"datetime"},
  {"name":"顾问角色","type":"select","multiple":false,"options":[{"name":"财务"},{"name":"营销"},{"name":"增长"},{"name":"运营"},{"name":"内容策略"},{"name":"关系BD"},{"name":"竞争行业"},{"name":"综合风险"}]},
  {"name":"分析摘要","type":"text"},
  {"name":"建议列表","type":"text"},
  {"name":"数据来源","type":"text"},
  {"name":"优先级","type":"number"}
]'
lark-cli base +table-create --name "汇总建议" --fields '[
  {"name":"建议编号","type":"text"},
  {"name":"建议内容","type":"text"},
  {"name":"来源角色","type":"text"},
  {"name":"优先级","type":"number"},
  {"name":"用户反馈","type":"select","multiple":false,"options":[{"name":"未读"},{"name":"已读"},{"name":"已展开"},{"name":"已执行"},{"name":"忽略"}]},
  {"name":"展开详情","type":"text"},
  {"name":"分析日期","type":"datetime"},
  {"name":"报告链接","type":"link"}
]'

Rate Limiting Handling: If +table-create partially fails due to rate limiting, the table will still be created successfully. In this case:

  1. Use base +table-list to get the table_id
  2. Use base +field-create --table-id <id> --json '{"name":"字段名","type":"text"}' to add fields one by one, with an interval of 1-2 seconds each time

Step 1: Data Collection

Based on the initialized Skills, pull the corresponding data:

# CRM 数据（如果 personal-crm 已初始化）
lark-cli base +record-list --table-id "<crm_contacts_table_id>" --filter '...热度分 < 30'
lark-cli base +record-list --table-id "<crm_interactions_table_id>" --filter '...互动时间 >= "7天前"'

# 社交追踪数据（如果 social-tracker 已初始化）
lark-cli base +record-list --table-id "<social_snapshot_table_id>" --filter '...日期 >= "7天前"'

# 创意/内容数据（如果 content-pipeline 已初始化）
lark-cli base +record-list --table-id "<idea_table_id>" --filter '...状态 != "已完成"'

# 投资数据（如果 stock-analyzer 已初始化）
lark-cli base +record-list --table-id "<portfolio_table_id>"
lark-cli base +record-list --table-id "<analysis_table_id>" --filter '...分析日期 >= "7天前"'

# 任务和日程（始终可用）
lark-cli task +get-my-tasks --page-all
lark-cli calendar +agenda --start "7天前" --end "今天"

Note: If the data table for an uninitialized Skill does not exist, simply skip that data source; it will not affect other analyses.

Step 2: Independent Analysis by 8 Roles

1. Financial Advisor

Data of interest: Investment portfolio (stock-analyzer), income/expense trends Analysis content:

  • Investment portfolio performance (profit/loss, volatility)
  • Income trend changes
  • Whether expenses are reasonable
  • Cash flow health

2. Marketing Advisor

Data of interest: Social media data (social-tracker), content performance Analysis content:

  • Interaction trends across platforms
  • Which content types perform best
  • Whether the content publishing cadence is reasonable
  • Audience growth/churn trends

3. Growth Advisor

Data of interest: Follower growth, channel customer acquisition Analysis content:

  • Overall growth trends
  • Fastest/slowest growing channels
  • Changes in growth rate
  • Comparison with peers

4. Operations Advisor

Data of interest: Task completion rate, schedule density Analysis content:

  • Task completion rate (completed/total tasks)
  • Whether the schedule is reasonable (overloaded/idle)
  • System usage efficiency
  • Process bottlenecks

5. Content Strategy Advisor

Data of interest: Idea library (content-pipeline), content performance Analysis content:

  • Distribution of topic quality
  • Idea execution progress
  • Cross-platform content performance comparison
  • Content gaps (which areas are not yet covered)

6. Relationship/BD Advisor

Data of interest: CRM contacts (personal-crm), meeting records Analysis content:

  • Cold contacts (need follow-up)
  • New contact trends
  • Relationship depth distribution
  • Identification of collaboration opportunities

7. Competition/Industry Advisor

Data of interest: Knowledge base (knowledge-base), industry trends Analysis content:

  • Industry trend changes
  • Competitor developments
  • New technologies/new models
  • New areas to watch

8. Comprehensive Risk Advisor

Data of interest: All data Analysis content:

  • Potential risks in each dimension
  • Identification of weak links
  • Issues that need priority handling
  • Warning signals

Step 3: Aggregate, Deduplicate, and Rank

The AI "Chief Secretary" role aggregates all advisors' suggestions:

  1. Deduplicate: Merge identical or highly similar suggestions raised by different advisors
  2. Priority ranking: By overall importance and urgency
  3. Numbering: Assign a number to each suggestion
  4. Category labeling: Label the domain to which the suggestion belongs

Output format:

## 今日商业顾问团报告（{日期}）

### 总览
今日共收到 {N} 条独立建议，来自 {M} 位顾问。
数据来源：CRM({有/无})、社交追踪({有/无})、创意库({有/无})、投资({有/无})

### 建议清单

1. 【财务】你的投资组合本周下跌 5%，主要受科技股拖累。建议检查 AAPL 的持仓比例是否过高。
2. 【关系】张三（XX科技）已 45 天未联系，上次讨论了合作意向，建议主动跟进。
3. 【内容】本周内容发布频率下降 40%，建议安排时间完成创意库中的待执行选题。
4. 【运营】你有 7 个任务已过期未完成，其中 3 个超过 14 天，建议集中清理。
5. 【风险】竞品 YY 产品本周发布新功能，与你的产品路线图有重叠，建议关注。
...

Step 4: Send and archive

Send to Feishu

lark-cli im +messages-send --chat-id "<chat_id>" --markdown "<清单内容>"

Archive report

lark-cli docs +create --title "商业顾问团日报 ({日期})" --markdown "<完整报告>"

Save analysis record

lark-cli base +record-upsert --table-id "<analysis_table_id>" --json '{"字段名":"值"}'
lark-cli base +record-upsert --table-id "<summary_table_id>" --json '{"字段名":"值"}'

Mode 2: Expand detailed analysis

When the user replies with a number (e.g., "tell me more about item 3"):

  1. Find the suggestion and source role corresponding to that number
  2. Re-analyze the role's complete viewpoint
  3. Provide more detailed arguments and data support
  4. Give specific action steps

Expansion format

## 第 {N} 条展开分析

### 原始建议
{原始建议内容}
### 来源顾问：{角色名}

### 详细分析
{展开的详细分析内容}

### 数据支撑
{具体数据和来源}

### 建议行动步骤
1. {步骤1}
2. {步骤2}
3. {步骤3}

### 预期效果
{如果执行该建议，预期会有什么改善}

Mode 3: Scheduled execution

Use CronCreate to set up automatic execution every night:

# 每晚 22:00 自动生成商业分析报告
CronCreate: cron="0 22 * * 1-5", prompt="运行商业顾问团分析，生成今日报告并发送到飞书", recurring=true

Permission table

Command Required scope
base +table-create bitable:app
base +record-list bitable:app:read
base +record-upsert bitable:app:write
docs +create docx:document:create
im +messages-send im:message:send_as_bot
task +get-my-tasks task:task:read
calendar +agenda calendar:calendar.event:read

Reference