pigdoc

Original：🇨🇳 Chinese

Translated

Search PIG official documentation and summarize answers. It is triggered when users ask technical questions, configuration methods, function descriptions related to PIG CLOUD, PIG AI, pigx, or explicitly say "check documentation", "search", "how to configure", "how to implement". When encountering specific problems related to the PIG framework, you should actively search the documentation before answering, and do not rely on potentially outdated information in the training data.

3installs

Sourcelltx/skills

Added on2026-04-10

NPX Install

npx skill4agent add lltx/skills pigdoc

SKILL.md Content (Chinese)

View Translation Comparison →

PIG Documentation Search Skill

The PIG official documentation library (

pig

index) contains the latest user guides, configuration instructions and best practices. Summarizing after searching can provide accurate official answers.

Search Steps

1. Extract Keywords

Extract 1-3 core keywords from the user's question. Chinese is preferred (the documentation is mainly in Chinese).

Examples:

"How to configure multiple data sources" →
```
多数据源
```
"OAuth2 login failure" →
```
OAuth2 登录
```
"Gateway routing 404 error" →
```
网关路由
```

2. Call the Search API

bash

curl -s -X POST "https://search.pig4cloud.com/indexes/pig/search" \
  -H "Content-Type: application/json" \
  -d '{"q": "关键词", "limit": 3}'

3. Response Field Description

Field	Meaning
`hits[].lvl0`	Document title
`hits[].content`	Document body (Markdown format)
`hits[].url`	Document path (e.g. `/doc/pigx/use/xxx` )
`hits[].lvl2`	Document update time

4. Summarize and Answer

Output in the following format based on search results:

According to PIG official documentation, regarding "[user question]":

Solution: [Synthesize the content of the documentation, extract core steps or descriptions, and list them in clear points. Retain key content such as code snippets and configuration examples]

Reference Documentation:

Document Title — A one-sentence description of the purpose of this document

Summarization Principles:

Retain code blocks, XML/YAML configurations, and key precautions in the documentation
Merge content when multiple documents describe the same issue, avoid duplication
If the relevance of search results to the question is low, explain it first and then supplement the answer based on known knowledge
If
```
hits
```
is empty, inform the user that there is no relevant content in the documentation for now

When Search Results Are Unsatisfactory

When the first search returns fewer than 2 hits or the content is obviously irrelevant, search again with another keyword, and summarize the combined results of the two searches.