/**
* Run SEO Audit Tool Handler
* Phase 2: Full implementation with crawling engine
*/
import { randomUUID } from 'crypto';
import path from 'path';
import os from 'os';
import { RunSeoAuditInputSchema } from '../schema/index.js';
import type { RunSeoAuditInput, RunSeoAuditOutput, CrawlConfig } from '../types/index.js';
import { UrlManager } from '../core/UrlManager.js';
import { ContentExtractor } from '../core/ContentExtractor.js';
import { LinkExtractor } from '../core/LinkExtractor.js';
import { CrawlStorage } from '../core/CrawlStorage.js';
import { CrawlOrchestrator } from '../core/CrawlOrchestrator.js';
import { debug } from '../utils/debug.js';
export async function runSeoAudit(params: RunSeoAuditInput): Promise<RunSeoAuditOutput> {
// DEBUG: Log raw incoming parameters
debug('[MCP] Raw params received:', JSON.stringify(params, null, 2));
const validated = RunSeoAuditInputSchema.parse(params);
// DEBUG: Log validated parameters
debug('[MCP] Validated params:', JSON.stringify(validated, null, 2));
const crawlId = randomUUID();
// Extract hostname for folder naming
const hostname = new URL(validated.url).hostname.replace(/^www\./, '');
const timestamp = new Date().toISOString().replace(/[:.]/g, '-').slice(0, 19);
const folderName = `${hostname}_${timestamp}_${crawlId.slice(0, 8)}`;
// Cross-platform output directory resolution
// Priority: OUTPUT_DIR env var > home directory fallback
const baseDir = process.env.OUTPUT_DIR || path.join(os.homedir(), 'seo-audits');
const outputPath = path.join(baseDir, folderName);
debug('[MCP] Output path resolved:', outputPath);
const config: CrawlConfig = {
crawlId,
startUrl: validated.url,
maxDepth: validated.depth ?? 3,
maxPages: validated.maxPages ?? 1000,
userAgent: validated.userAgent ?? 'chrome',
crawlExternal: false,
respectRobots: true,
concurrency: 5,
delay: 1000,
timeout: 30000,
includeExtensions: ['html', 'htm', 'php', 'asp', 'aspx', 'jsp'],
excludeExtensions: ['pdf', 'doc', 'zip', 'exe', 'jpg', 'png', 'gif', 'css', 'js', 'xml', 'rss', 'atom', 'json'],
includePatterns: [],
excludePatterns: [],
outputPath,
createdAt: new Date().toISOString()
};
// DEBUG: Log final config being passed to orchestrator
debug('[MCP] Final crawl config:', JSON.stringify(config, null, 2));
const storage = new CrawlStorage(outputPath, validated.url);
await storage.initialize();
await storage.saveConfig(config);
const baseDomain = new URL(config.startUrl).hostname;
const urlManager = new UrlManager(baseDomain);
const contentExtractor = new ContentExtractor();
const linkExtractor = new LinkExtractor(baseDomain);
const orchestrator = new CrawlOrchestrator(
config,
urlManager,
contentExtractor,
linkExtractor,
storage
);
const metadata = await orchestrator.run();
// Generate CSV export
await storage.generateCsvExport();
return {
crawlId: config.crawlId,
outputPath: config.outputPath,
status: metadata.status,
stats: metadata.stats
};
}
