Meeting Agenda Generator

Purpose

Generate professional Chinese meeting agenda documents in Word (.docx) and PDF formats.

Apply the standard font and layout conventions described in references/template.md.

Trigger Conditions

Use this skill when the user:

• Provides meeting information (title, time, location, agenda, participants) and asks to generate an agenda
• Mentions keywords like "会议议程"、"议程安排"、"生成议程"、"帮我做个会议议程"
• Provides raw meeting details and asks for a formatted document output

Workflow

Step 1: Collect Meeting Information

Ask the user to provide (or extract from their message):

• 会议名称 (Meeting title)
• 环节列表 (Session list), for each session:
• 环节名称 (Session name)
• 时间 (Time, e.g. "2026年4月2日（星期四）09:00-10:00")
• 地点 (Location)
• 内容 (Content description, optional for simple sessions)
• 主持人 (Host, optional)
• 议程 (Agenda items list, optional)
• 参会人员 (Participants by organization, optional)
• 单位名称 (Organization name)
• 成员列表: 姓名 + 职务 + 补充说明(optional)

If any critical fields are missing, ask the user before proceeding.

Use references/template.md for exact field names, structure, and JSON format.

Step 2: Build the JSON Data File

Construct a JSON file following the schema in references/template.md.

Save it as a temporary file, e.g. agenda_input.json, in a working directory

(suggest using the user's workspace or a temp folder).

Example minimal JSON:

{
  "title": "XXX合作座谈会议议程",
  "filename": "XXX合作座谈会议议程",
  "sections": [
    {
      "name": "产品演示",
      "time": "2026年4月2日（星期四）09:00-10:00",
      "location": "XX楼会议室",
      "content": "演示核心功能"
    }
  ]
}

Step 3: Run the Generation Script

Execute scripts/generate_agenda.py with the JSON data file:

python <skill_base_dir>/scripts/generate_agenda.py \
  --data <path_to_agenda_input.json> \
  --output <output_directory>

• is the absolute path to this skill directory

(typically ~/.workbuddy/skills/meeting-agenda)

• is where the files will be saved

(use the user's workspace directory or a subdirectory they specify)

The script will:

1. Auto-install python-docx and reportlab if not present
2. Generate .docx (Word document)
3. Generate .pdf (PDF document)

Step 4: Deliver Results

After generation:

1. Confirm both files exist
2. Report the file paths to the user
3. Use deliver_attachments tool to attach both .docx and .pdf files

Font & Style Reference

Refer to references/template.md for the complete font specification:

• Meeting title: FangSong GB2312 (仿宋_GB2312), 24pt (小一), bold, centered
• Section headings (including 人员): Heiti (黑体), 16pt, bold, left-aligned; numbered with Chinese numerals (一、二、三…)
• Field labels (时间/地点/内容/议程 etc.): Heiti (黑体), 16pt, bold
• Body text (field values, agenda items): FangSong GB2312 (仿宋_GB2312), 16pt
• Organization unit headers: Kaiti GB2312 (楷体_GB2312), 16pt, bold, preceded by two full-width spaces (　　)
• Member lines: FangSong GB2312, 16pt, NO leading dash; name and title separated by two full-width spaces (　　)

人员排列规则：

• 每个 section 如含 persons 字段，人员独立作为一个带序号的一级标题（如"三、人员"），序号接续前面环节编号

Error Handling

• If PDF generation fails (missing fonts or reportlab issues), still deliver the Word file and explain the PDF issue
• If the user's system lacks Chinese fonts, the Word file will still display correctly when opened in Word; the PDF may show fallback fonts
• If python-docx install fails, ask the user to run pip install python-docx reportlab manually