import { z } from 'zod/v3';
// Search Schema
export const SearchSchema = z.object({
query: z.string().describe('Search query string'),
limit: z.number().optional().describe('Maximum number of results to return (default: 10)'),
language: z.string().optional().describe('Language code for search results (default: auto)'),
categories: z.enum([
'general',
'news',
'images',
'videos',
'it',
'science',
'map',
'music',
'files',
'social_media',
]).optional().describe('Categories to search for (default: general)'),
timeRange: z.enum([
'all',
'day',
'week',
'month',
'year',
]).optional().describe('Time range for search results (default: all)'),
});
export type SearchInput = z.infer<typeof SearchSchema>;
// Map Schema
export const MapSchema = z.object({
url: z.string().describe('Starting URL for URL discovery'),
search: z.string().optional().describe('Optional search term to filter URLs'),
ignoreSitemap: z.boolean().optional().describe('Skip sitemap.xml discovery and only use HTML links'),
sitemapOnly: z.boolean().optional().describe('Only use sitemap.xml for discovery, ignore HTML links'),
includeSubdomains: z.boolean().optional().describe('Include URLs from subdomains in results'),
limit: z.number().optional().describe('Maximum number of URLs to return'),
});
export type MapInput = z.infer<typeof MapSchema>;
// Action Schema for Scrape
const ActionSchema = z.discriminatedUnion('type', [
z.object({
type: z.literal('wait'),
milliseconds: z.number().describe('Time to wait in milliseconds'),
}),
z.object({
type: z.literal('click'),
selector: z.string().describe('CSS selector for the target element'),
}),
z.object({
type: z.literal('screenshot'),
fullPage: z.boolean().optional().describe('Take full page screenshot'),
}),
z.object({
type: z.literal('write'),
selector: z.string().describe('CSS selector for the target element'),
text: z.string().describe('Text to write'),
}),
z.object({
type: z.literal('press'),
key: z.string().describe('Key to press'),
}),
z.object({
type: z.literal('scroll'),
direction: z.enum(['up', 'down']).describe('Scroll direction'),
}),
z.object({
type: z.literal('scrape'),
}),
z.object({
type: z.literal('executeJavascript'),
script: z.string().describe('JavaScript code to execute'),
}),
]);
// Scrape Schema
export const ScrapeSchema = z.object({
url: z.string().describe('The URL to scrape'),
formats: z.array(z.enum([
'markdown',
'html',
'rawHtml',
'screenshot',
'links',
'screenshot@fullPage',
'extract',
])).optional().describe("Content formats to extract (default: ['markdown'])"),
onlyMainContent: z.boolean().optional().describe('Extract only the main content, filtering out navigation, footers, etc.'),
includeTags: z.array(z.string()).optional().describe('HTML tags to specifically include in extraction'),
excludeTags: z.array(z.string()).optional().describe('HTML tags to exclude from extraction'),
waitFor: z.number().optional().describe('Time in milliseconds to wait for dynamic content to load'),
timeout: z.number().optional().describe('Maximum time in milliseconds to wait for the page to load'),
actions: z.array(ActionSchema).optional().describe('List of actions to perform before scraping'),
extract: z.object({
schema: z.record(z.any()).optional().describe('Schema for structured data extraction'),
systemPrompt: z.string().optional().describe('System prompt for LLM extraction'),
prompt: z.string().optional().describe('User prompt for LLM extraction'),
}).optional().describe('Configuration for structured data extraction'),
mobile: z.boolean().optional().describe('Use mobile viewport'),
skipTlsVerification: z.boolean().optional().describe('Skip TLS certificate verification'),
removeBase64Images: z.boolean().optional().describe('Remove base64 encoded images from output'),
location: z.object({
country: z.string().optional().describe('Country code for geolocation'),
languages: z.array(z.string()).optional().describe('Language codes for content'),
}).optional().describe('Location settings for scraping'),
});
export type ScrapeInput = z.infer<typeof ScrapeSchema>;
// Extract Schema
export const ExtractSchema = z.object({
urls: z.array(z.string()).describe('List of URLs to extract information from'),
prompt: z.string().optional().describe('Prompt for the LLM extraction'),
systemPrompt: z.string().optional().describe('System prompt for LLM extraction'),
schema: z.record(z.any()).optional().describe('JSON schema for structured data extraction'),
allowExternalLinks: z.boolean().optional().describe('Allow extraction from external links'),
enableWebSearch: z.boolean().optional().describe('Enable web search for additional context'),
includeSubdomains: z.boolean().optional().describe('Include subdomains in extraction'),
});
export type ExtractInput = z.infer<typeof ExtractSchema>;
