zai-search
by bartoszpijet
Web & Frontend DevelopmentBrief
AI-optimized web search via Z.AI Web Search API.
Description
What This Skill Does
Web search tool built on the Z.AI Web Search API, returning structured results with titles, URLs, and summaries formatted for LLM consumption. Supports filtering by result count, domain, and time range. Includes basic URL content extraction via native fetch.
Returns pre-structured results with metadata fields LLMs can directly consume, skipping manual HTML parsing or scraping.
When to Use It
- Finding recent news on a topic for an AI agent pipeline
- Restricting search results to a single domain for targeted research
- Extracting readable text from a specific URL
- Filtering results by time range to get content from the past day, week, or month
- Supplying structured search context to an LLM before generating a response
Example Workflow
User asks: find articles about climate policy published in the last week
- 1
Run search.mjs with the query and --recency oneWeek flag
- 2
Receive structured results containing titles, URLs, and summaries
- 3
Pass the results array to the LLM as context
- 4
LLM synthesizes a response from the provided summaries
A list of recent climate policy articles with titles, URLs, and summaries ready for LLM processing
Requirements
- ZAI_API_KEY environment variable, obtained by signing up at chat.z.ai
Installation
clawhub install bartoszpijet/zai-searchRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.