agentsclimarketplace

Event collection

Skill Makiato1999/agentic-travel/.claude/skills/event-collection

An LLM-powered multi-agent business travel planning system that integrates RAG knowledge retrieval, user preference memory, real-time information querying, and itinerary generation.

Install
npx -y skills add Makiato1999/agentic-travel --skill event-collection

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use this skill when the user provides travel details like origin, destination, dates, purpose, or when planning a trip. It extracts structured event information for itinerary planning.

SKILL.md

3.0 KB, 980 tokens by cl100k_base, as published. Nobody here has run it

Event Collection Skill

用于从用户输入中提取本次行程的结构化信息,供 itinerary_planning 等后续 agent 使用。

通常由 IntentionAgent 自动调度,配合 plan-trip 技能使用。

核心原则

  1. 只提取当前这一次行程的信息,不要把用户历史偏好、家庭住址、常住地当作本次行程事实
  2. 如果用户输入里明确出现了出发地、目的地、日期、时长,必须以当前输入中的显式信息为准,优先级高于背景信息
  3. 只有当用户当前输入没有明确说出发地时,才允许根据背景信息中的 home_location 推断 origin
  4. 用户说“常住苏州”“我家在杭州”这类信息,默认属于背景信息/偏好,不等于本次出发地,除非用户明确说“从苏州出发”
  5. 如果用户当前输入中明确说“从南京去成都”,则 origin 必须是南京,不能被家庭住址或历史偏好覆盖

提取字段

请尽可能提取以下结构化字段:

  1. origin - 出发地
  2. destination - 目的地
  3. start_date - 出发日期(YYYY-MM-DD
  4. end_date - 返程日期(YYYY-MM-DD
  5. duration_days - 行程天数
  6. return_location - 返程地
  7. trip_purpose - 行程目的
  8. missing_info - 缺失字段列表
  9. extracted_count - 成功提取字段数量
  10. summary - 一句简要摘要

日期处理规则

  • 当前时间由调用方提供
  • 用户说“2月27日”“2.27”等相对日期时,需要根据当前时间推断完整年月日
  • 用户说“明天”“后天”“下周”等相对时间时,需要换算成具体日期
  • 所有日期必须输出为完整的 YYYY-MM-DD
  • 如果用户给了 duration_days = N 且给了 start_date,但没有明确 end_date,则按包含起始日计算:
    • 1天 -> end_date = start_date
    • 2天 -> end_date = start_date + 1天
    • 4天 -> end_date = start_date + 3天
    • 一般规则:end_date = start_date + (duration_days - 1)

特殊处理

  • 对于“北京一日游”这类:origindestination 都设为北京
  • 对于“一日游”:duration_days = 1
  • 如果用户没明确说 return_location,默认 return_location = origin
  • “常住地 / 家在某地 / 喜欢某类酒店 / 饮食习惯”不属于 trip_purpose,不要混入 origin/destination/trip_purpose

输出格式

严格输出 JSON:

{
  "origin": "北京",
  "destination": "上海",
  "start_date": "2026-04-13",
  "end_date": "2026-04-16",
  "duration_days": 4,
  "return_location": "北京",
  "trip_purpose": "商务出行",
  "missing_info": [],
  "extracted_count": 7,
  "summary": "2026年4月13日至16日从北京到上海的4天商务出行"
}

缺失的信息在 missing_info 中列出,对应字段设为 null

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.