AgentSkillsCN

xmtp-docs

查询XMTP文档,了解当前SDK的模式、方法与示例。当您需要:(1) 查找XMTP SDK的方法或模式;(2) 在编码前验证当前API签名;(3) 寻找XMTP功能的示例,如流媒体、同意机制、群组或内容类型时,可使用此功能。

SKILL.md
--- frontmatter
name: xmtp-docs
description: >
  Query XMTP documentation for current SDK patterns, methods, and examples.
  Use when: (1) Looking up XMTP SDK methods or patterns, (2) Verifying
  current API signatures before coding, (3) Finding examples for XMTP
  features like streaming, consent, groups, or content types.

XMTP Documentation Lookup

Query current XMTP documentation to find accurate SDK patterns before writing code.

Why This Skill Exists

CRITICAL: Never use training data for XMTP SDK methods. The SDK evolves frequently and method names change between versions. All method names, signatures, and patterns MUST be looked up from current documentation.

How to Query

Step 1: Find the right page

Use the docs index to find which page has what you need:

code
WebFetch({
  url: "https://docs.xmtp.org/llms.txt",
  prompt: "Find the page URL for [topic]"
})

Step 2: Fetch that page

Fetch the specific page for complete code examples:

code
WebFetch({
  url: "https://docs.xmtp.org[path-from-step-1]",
  prompt: "Extract [specific feature] with code examples"
})

Examples

Find how to create a client:

code
WebFetch({ url: "https://docs.xmtp.org/llms.txt", prompt: "Find the page URL for creating an XMTP client" })
// Returns: /chat-apps/core-messaging/create-a-client

WebFetch({ url: "https://docs.xmtp.org/chat-apps/core-messaging/create-a-client", prompt: "Extract how to create a client with code examples" })

Find how to stream messages:

code
WebFetch({ url: "https://docs.xmtp.org/llms.txt", prompt: "Find the page URL for streaming messages" })
// Returns: /chat-apps/list-stream-sync/stream

WebFetch({ url: "https://docs.xmtp.org/chat-apps/list-stream-sync/stream", prompt: "Extract streaming patterns with code examples" })

Tips

  1. Always use the index first - Don't guess URLs; they change when docs reorganize
  2. Ask for code examples - Docs include examples for Browser, Node, Kotlin, and Swift
  3. One topic per query - Focused queries return better results
  4. Don't use llms-full.txt - It's 500KB+ and WebFetch can't reliably search it