x jina serves as the "eyes" and "external brain" for AI Agents. Its core strength lies in converting complex web pages into AI-friendly Markdown and providing powerful search and semantic ranking capabilities.
When to Activate
- When reading and analyzing the content of a specific web page (URL), automatically converted to Markdown.
- When performing real-time web searches to gather the latest information.
- When identifying the most semantically relevant segments from a set of documents (Reranker).
Core Principles & Rules
- Markdown First: Retrieves web pages in Markdown by default, which is optimal for LLM consumption.
- Environment Requirements: Requires a Jina API Key. If not configured, AI should guide the user through initialization.
- Configuration Guidance:
- Direct the user to jina.ai to obtain a free or paid API Key.
- Suggest the user run
x jina initfor configuration.
Patterns & Examples
Read Web Content (Most Common)
# Retrieve web content and convert to Markdown for AI reading
x jina reader https://example.com/article
Real-time Web Search
# Search for keywords and return summaries of the top 5 results
x jina search "latest AI trends 2024"
Semantic Reranking
# Find the top 3 segments from a file most relevant to "how to install"
x jina reranker generate -f docs.txt --top 3 "how to install"
Text Embedding
# Generate vector data for text, useful for vector database retrieval
x jina embed generate "This is a test text"
Configuration Guide (for AI)
If an API error occurs, provide this guidance to the user:
Please obtain an API Key from Jina AI (https://jina.ai), then run the following command in your terminal to initialize:
x jina init
Checklist
- Prioritize using the
readersubcommand for Markdown. - Combine search tasks with
rerankerfor better precision.