Mikan RSS 番剧订阅 Skill

This repository provides a Cursor Agent Skill for fetching anime RSS subscription links and episode lists from mikanani.kas.pub (蜜柑计划).

What this skill does

• Normalize user requests about Mikan shows:
  • Understand year + season (e.g. 2025 秋季番, 2026 冬季番)
  • Parse show titles and optional subtitle group preferences
• Locate bangumi by either season or keyword search:
  • Use the seasonal lineup when the user 给出了 年份 + 季度
  • Use the 搜索结果页 (/Home/Search?searchstr=) when the user只给番名或关键字，再在需要时回退到季度检索
• Scrape bangumi detail pages to:
  • List available subtitle groups and their IDs
  • Build per‑group RSS links
  • Optionally list recent episodes (file name, size, updated time, episode page)

Typical use cases

• Get RSS links for all groups for a given show and season
  e.g. “2026 年冬季《优雅贵族的休假指南。》所有字幕组的 RSS”
• Get RSS only for a specific group
  e.g. “只要 LoliHouse 的订阅链接”
• Explore a season’s entire lineup to find the correct bangumi and its groups
• Only provide a bangumi title (or keywords) and let the skill search Mikan, resolve to a specific 番剧，并给出各字幕组的订阅链接

High‑level workflow

1. Parse user intent
   • Extract year, season, bangumi title, and (optional) subgroup name
   • If season/year are omitted, first try the keyword search page to locate the correct 番剧；若无法唯一确定，再按推断的 year + season 去季度列表中查找
2. Fetch candidate bangumi list
   • From either:
     • the seasonal lineup: https://mikanani.kas.pub/Home/BangumiCoverFlowByDayOfWeek?year=<YEAR>&seasonStr=<SEASON>, or
     • the search page: https://mikanani.kas.pub/Home/Search?searchstr=<ENCODED_KEYWORD> (只解析主体列表区域)
   • Collect all bangumi entries (bangumiId, title, etc.) and match the requested title (exact match first, then fuzzy)
3. Fetch bangumi detail
   • Visit https://mikanani.kas.pub/Home/Bangumi/<bangumiId>
   • Extract all subtitle groups and their IDs
   • For each group, build RSS:
     • https://mikanani.kas.pub/RSS/Bangumi?bangumiId=<BANGUMI_ID>&subgroupid=<SUBGROUP_ID>
   • Optionally parse recent episodes for that group
4. Filter and format results
   • If the user specified a group, filter to that group; otherwise list all
   • Clearly show: bangumi info, each subtitle group, its RSS URL, and (optionally) latest few episodes
   • Mention any assumptions about season/year or ambiguous title matches, and list 候选番剧供用户确认

Repository structure

• SKILL.md – Full, detailed skill specification and examples for the Cursor agent (包括季度接口与搜索接口的解析细节)。
• README.md – High‑level overview (this file).

Notes

• This skill is focused on Mikan RSS and resource discovery, not on video playback or general streaming.
• When unsure about the user’s exact show or season, the agent should:
  • Make a reasonable assumption,
  • Clearly state it,
  • Provide alternatives when multiple close matches exist, especially 在仅通过搜索关键字无法唯一定位番剧时。

Mikan RSS 番剧订阅 Skill

This repository provides a Cursor Agent Skill for fetching anime RSS subscription links and episode lists from mikanani.kas.pub (蜜柑计划).

What this skill does

• Normalize user requests about Mikan shows:
  • Understand year + season (e.g. 2025 秋季番, 2026 冬季番)
  • Parse show titles and optional subtitle group preferences
• Fetch season lineups from Mikan and locate a specific bangumi by name
• Scrape bangumi detail pages to:
  • List available subtitle groups
  • Build per‑group RSS links
  • Optionally list recent episodes (file name, size, updated time, episode page)

Typical use cases

• Get RSS links for all groups for a given show and season
  e.g. “2026 年冬季《优雅贵族的休假指南。》所有字幕组的 RSS”
• Get RSS only for a specific group
  e.g. “只要 LoliHouse 的订阅链接”
• Explore a season’s entire lineup to find the correct bangumi and its groups

High‑level workflow

1. Parse user intent
   • Extract year, season, bangumi title, and (optional) subgroup name
   • If season/year are omitted, assume the most relevant current/previous season and state that assumption
2. Fetch seasonal bangumi list
   • Call https://mikanani.kas.pub/Home/BangumiCoverFlowByDayOfWeek?year=<YEAR>&seasonStr=<SEASON>
   • Collect all Bangumi entries (bangumiId, title, etc.)
   • Match the requested title (exact match first, then fuzzy)
3. Fetch bangumi detail
   • Visit https://mikanani.kas.pub/Home/Bangumi/<bangumiId>
   • Extract all subtitle groups and their IDs
   • For each group, build RSS:
     • https://mikanani.kas.pub/RSS/Bangumi?bangumiId=<BANGUMI_ID>&subgroupid=<SUBGROUP_ID>
   • Optionally parse recent episodes for that group
4. Filter and format results
   • If the user specified a group, filter to that group; otherwise list all
   • Clearly show: bangumi info, each subtitle group, its RSS URL, and (optionally) latest few episodes
   • Mention any assumptions about season/year or ambiguous title matches

Repository structure

• SKILL.md – Full, detailed skill specification and examples for the Cursor agent.
• README.md – High‑level overview (this file).

Notes

• This skill is focused on Mikan RSS and resource discovery, not on video playback or general streaming.
• When unsure about the user’s exact show or season, the agent should:
  • Make a reasonable assumption,
  • Clearly state it,
  • And provide alternatives when multiple close matches exist.