lx-url-route

URL routing

Prerequisite: Requires the lx CLI to be configured and logged in.

⚡ When to use this skill?

Entry scenarios

• The user only sent a link and did not explicitly say whether they want to view, edit, import, or comment
• Need to first determine whether this is a page, knowledge base, team, or official account article link

Prohibited in this skill

• Do not route scenarios with a clear intent: the user has already said they want to "edit the page"/"view comments"/"create a page" → switch directly to the corresponding skill
• Do not show internal MCP domains to the user: when generating user-visible links, only use lexiang.tencent.com

⚡ How to choose a command? (Decision tree)

收到 URL →
├── /pages/{entry_id}        → 默认 `lx entry describe-ai-parse-content`
├── /spaces/{space_id}       → `lx space describe-space`
├── /t/{team_id}/spaces      → `lx space list-spaces`
├── mp.weixin.qq.com/*       → 先补齐目标位置，再 `lx file create-hyperlink`
├── lexiang.tencent.com/ai/claw → Token 配置引导，不调用命令
├── mcp.lexiang-app.com/*    → 内部端点，不展示给用户
└── 其他                     → 告知不是可处理的乐享链接

⚠️ High-risk operations and default preferred paths

Default behavior:

• The user only sent a page link and did not say what to do → default to viewing the content, use lx entry describe-ai-parse-content
• The user only sent a knowledge base link → first use lx space describe-space to get the root_entry_id, then decide whether to switch to lx-entry
• The user only sent a team link → first list the knowledge bases under that team, do not directly guess the target knowledge base

Information that must be completed:

• Before importing a WeChat official account article, you must first confirm space_id and parent_entry_id
• If the user wants to edit page content, view comments, or create a page, do not stay in this skill; immediately switch to the downstream skill

URL types and default actions

URL pattern	Extracted fields	Default action	Destination when user intent is clear
/pages/{entry_id}	entry_id	lx entry describe-ai-parse-content --entry-id <ID>	View details → lx entry describe-entry; edit content → lx-block; view comments → lx-connector
/spaces/{space_id}	space_id	lx space describe-space --space-id <ID>	Browse directory / create page → lx-space → lx-entry
/t/{team_id}/spaces	team_id	lx space list-spaces --team-id <ID>	After selecting a knowledge base, continue with lx-space / lx-entry
mp.weixin.qq.com/*	Original URL	First complete the target location, then lx file create-hyperlink --url ... --space-id ... --parent-entry-id ...	After import is complete, if the page needs to be edited, then switch to lx-entry / lx-block
lexiang.tencent.com/ai/claw	None	Prompt the user to go to that page to obtain the Token	Do not invoke a command
mcp.lexiang-app.com/*	None	Internal service endpoint	Do not show to the user

🎯 Execution rules

1. First determine whether routing is needed: use this skill only when "the user provided a URL, but the intent is unclear".
2. Page links default to viewing content: the default action is describe-ai-parse-content, not describe-entry。
3. Importing a WeChat official account article must first complete the target location: when space_id / parent_entry_id is missing, you cannot import directly.
4. Do not expose internal domains：mcp.lexiang-app.com is only used for internal service identification and must not be echoed back to the user.
5. Exit once routing is complete: after obtaining the target ID or determining the downstream command, you should switch to the corresponding skill and not continue performing complex operations within this skill.