
12306
Query China Railway 12306 for train schedules, remaining tickets, and station info. Use when user asks about train/高铁/火车 tickets, schedules, or availability within China.
- Installs
- 640
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/kirorab/12306-skill --skill 12306- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads 12306 automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
12306 Train Query
Query train schedules and remaining tickets from China Railway 12306.
Query Tickets
node {baseDir}/scripts/query.mjs <from> <to> [options]
- HTML mode (default): writes file, prints path to stdout
- Markdown mode (
-f md): prints table to stdout
Examples
# All trains from Beijing to Shanghai (defaults to today)
node {baseDir}/scripts/query.mjs 北京 上海
# Markdown table output (to stdout, good for chat)
node {baseDir}/scripts/query.mjs 北京 上海 -t G -f md
# Morning departures, 2h max, with second class available
node {baseDir}/scripts/query.mjs 上海 杭州 -t G --depart 06:00-12:00 --max-duration 1h --seat ze
# Only bookable trains arriving before 6pm
node {baseDir}/scripts/query.mjs 深圳 长沙 --available --arrive -18:00
# Custom output path
node {baseDir}/scripts/query.mjs 广州 武汉 -o /tmp/tickets.html
# JSON output (to stdout)
node {baseDir}/scripts/query.mjs 广州 武汉 --json
Options
-d, --date <YYYY-MM-DD>: Travel date (default: today)-t, --type <G|D|Z|T|K>: Filter train types (combinable, e.g.GD)--depart <HH:MM-HH:MM>: Depart time range (e.g.08:00-12:00,18:00-)--arrive <HH:MM-HH:MM>: Arrive time range (e.g.-18:00,14:00-20:00)--max-duration <duration>: Max travel time (e.g.2h,90m,1h30m)--available: Only show bookable trains--seat <types>: Only show trains with tickets for given seat types (comma-separated:swz,zy,ze,rw,dw,yw,yz,wz)-f, --format <html|md>: Output format —html(default, saves file) ormd(markdown table to stdout)-o, --output <path>: Output file path, html mode only (default:{baseDir}/data/<from>-<to>-<date>.html)--json: Output raw JSON to stdout
Output Columns
| Column | Meaning |
|---|---|
| 商务/特等 | Business class / Premium (swz) |
| 一等座 | First class (zy) |
| 二等座 | Second class (ze) |
| 软卧/动卧 | Soft sleeper / Bullet sleeper (rw/dw) |
| 硬卧 | Hard sleeper (yw) |
| 硬座 | Hard seat (yz) |
| 无座 | Standing (wz) |
Values: number = remaining seats, 有 = available (qty unknown), — = not applicable
Station Lookup
node {baseDir}/scripts/stations.mjs 杭州
node {baseDir}/scripts/stations.mjs 香港西九龙
Important Notes for AI Assistant
⚠️ Station Name Resolution Warning
CRITICAL: When querying by city name (e.g., "武汉", "上海", "深圳", "广州"), the API may return trains from/to ANY station in that city, not just the main station.
Common Pitfalls:
- 武汉 includes: 武汉站 (main), 汉口站 (Hankou), 武昌站 (Wuchang), 武汉东站
- 上海 includes: 上海虹桥 (Hongqiao), 上海站 (main), 上海南站, 上海松江站
- 深圳 includes: 深圳北站 (main), 深圳站 (Luohu), 福田站, 深圳东站
- 广州 includes: 广州南站 (main), 广州站, 广州东站, 广州北站
Best Practice - Always verify exact stations:
- First, use
stations.mjsto list all stations in the city:node {baseDir}/scripts/stations.mjs 武汉 - Then, query with exact station names for accurate results:
node {baseDir}/scripts/query.mjs 武汉 上海虹桥 -f md
🔄 Transfer/Connection Guidelines
When planning transfers (中转):
- Use JSON output (
--json) to verify exact station names - Ensure both segments use the SAME station (e.g., both use 武汉站, not 武汉→汉口)
- Recommended minimum transfer time: 20-30 minutes for same station
- Different stations in same city require additional travel time (e.g., 武汉→汉口 = 30+ min by subway)
📋 Query Workflow Recommendation
For accurate results, follow this workflow:
-
List stations in departure city:
node {baseDir}/scripts/stations.mjs 北京 -
List stations in arrival city:
node {baseDir}/scripts/stations.mjs 上海 -
Query with exact station names (e.g., 北京南 → 上海虹桥):
node {baseDir}/scripts/query.mjs 北京南 上海虹桥 -d 2026-03-05 -f md -
For transfers: Always verify both segments use the same station by checking
fromStationandtoStationin JSON output.
Technical Notes
- Data comes directly from 12306 official API (no key needed)
- Station data is cached for 7 days in
{baseDir}/data/stations.json - Works for all train types: G (高铁), D (动车), Z (直达), T (特快), K (快速)
FAQ
What does the 12306 skill do?
Query China Railway 12306 for train schedules, remaining tickets, and station info. Use when user asks about train/高铁/火车 tickets, schedules, or availability within China.
How do I install 12306?
Run npx skills add https://github.com/kirorab/12306-skill --skill 12306 in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/12306/ (personal) or .claude/skills/12306/ (project).
Which AI agents can use 12306?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is 12306 safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains 12306?
It is published by kirorab in kirorab/12306-skill. It has 640 recorded installs.
Popular other skills
hyperframes-cli
heygen-com/hyperframes
The HyperFrames CLI by job - scaffold, check, look at frames, edit the timeline, preview, render locally or in the cloud, publish, project history - plus reading its JSON and recovering from build or render failures.
triage
mattpocock/skills
Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs.
ask-matt
mattpocock/skills
Ask which skill or flow fits your situation. A router over the skills in this repo.
to-questionnaire
mattpocock/skills
Turn a decision you can't fully answer into a questionnaire for someone else to fill in.
loop-me
mattpocock/skills
Grill me about specs for the workflows I want to build, within this workspace.