Node SDK Reference

import { KBClient } from 'sarvcrawl';

const kb = new KBClient({
  apiKey: 'YOUR_API_KEY',       // or set KB_API_KEY env var
  apiUrl: 'https://crawl.sarv.com', // the default; set your own host to self-host
});

const status = await kb.health();
console.log(status);

const user = await kb.me();
console.log(user.id, user.email);

const newKb = await kb.create('Research KB', {
  description: 'My research knowledge base',
  settings: {},
});
console.log(newKb.id); // "kb_abc123"

const kbs = await kb.list();
kbs.forEach(k => console.log(k.id, k.name));

const myKb = await kb.get('kb_abc123');
console.log(myKb.total_pages, myKb.total_mb);

const stats = await kb.stats('kb_abc123');
console.log(stats.total_pages, stats.total_mb);

await kb.deleteKb('kb_abc123');

const job = await kb.scrape('kb_abc123', 'https://example.com/article', {
  formats: ['markdown', 'html'],  // default: ['markdown', 'json']
  include_docs: true,             // auto-download linked PDFs/DOCX
});
console.log(job.job_id, job.status); // "queued"

const job = await kb.crawl('kb_abc123', 'https://docs.example.com', {
  formats: ['markdown', 'json'],
  max_depth: 5,
  limit: 500,
  exclude_paths: ['/admin', '/api'],
  include_docs: true,
  allow_backward_links: false,
});
console.log(job.job_id);

const job = await kb.search('kb_abc123', 'knowledge infrastructure 2024', {
  limit: 10,    // number of results to scrape, default: 10
  lang: 'en',   // result language filter, default: 'en'
  formats: ['markdown', 'json'],
});
console.log(job.job_id);

const job = await kb.map('kb_abc123', 'https://example.com', {
  limit: 5000,  // max URLs to discover, default: 5000
});
console.log(job.job_id);

const job = await kb.upload('kb_abc123', '/path/to/report.pdf', {
  formats: ['markdown', 'json'],
  ocr_language: 'eng',  // default: 'eng'. Also supports 'hin'
});
console.log(job.job_id);

const job = await kb.monitor('kb_abc123', 'page', {
  urls: ['https://example.com/pricing'],
  schedule: 'daily',
  notify_url: 'https://your-endpoint.example.com/webhook',
});
console.log(job.job_id, job.status); // "queued"

// Website mode — crawl and watch every discovered page:
await kb.monitor('kb_abc123', 'website', {
  url: 'https://docs.example.com',
  limit: 100,
  include_paths: ['/docs'],
  schedule: 'every 6 hours',
  notify_url: 'https://your-endpoint.example.com/webhook',
});

const job = await kb.job('dacf0540-f611-47eb-8e47-d5c72acea679');
console.log(job.status, job.urls_processed);

const jobs = await kb.jobs({
  kb_id: 'kb_abc123',   // filter by KB
  status: 'completed',  // queued | running | completed | failed | cancelled
  type: 'scrape',       // scrape | crawl | search | map | upload
  limit: 20,
  offset: 0,
});

await kb.cancelJob('dacf0540-f611-47eb-8e47-d5c72acea679');

await kb.deleteJob('dacf0540-f611-47eb-8e47-d5c72acea679');

const logs = await kb.jobLogs('dacf0540-f611-47eb-8e47-d5c72acea679');
logs.forEach(l => console.log(l.level, l.message));

const audit = await kb.jobAudit('dacf0540-f611-47eb-8e47-d5c72acea679');
console.log(audit.similarity_score);

const pages = await kb.pages('kb_abc123', {
  limit: 50,
  offset: 0,
});
pages.forEach(p => console.log(p.url, p.title));

const page = await kb.page('kb_abc123', 'page_abc', 'markdown');
console.log(page.content);

const results = await kb.searchKb('kb_abc123', 'knowledge infrastructure', {
  page: 1,
  size: 10,
  type: 'scrape',    // optional: filter by job type
  job_id: 'job_xyz', // optional: filter by specific job
});
results.hits.forEach(h => console.log(h.title, h.score));

const results = await kb.embedSearchKb('kb_abc123', 'how do refunds work?', {
  size: 10,
  type: 'crawl',     // optional: filter by job type
  job_id: 'job_xyz', // optional: filter by specific job
});
results.hits.forEach(h => console.log(h.rrf_score, h.title, h.text));

const jobs = await kb.kbJobs('kb_abc123', { limit: 20, offset: 0 });

const files = await kb.jobFiles('kb_abc123', 'dacf0540-f611-47eb-8e47-d5c72acea679');
files.forEach(f => console.log(f.path, f.size));

const buffer = await kb.downloadFile(
  'kb_abc123',
  'dacf0540-f611-47eb-8e47-d5c72acea679',
  'pages/page_abc/content.md',
);
fs.writeFileSync('content.md', buffer);

const zip = await kb.exportJobZip('kb_abc123', 'dacf0540-f611-47eb-8e47-d5c72acea679');
fs.writeFileSync('job-export.zip', zip);

const zip = await kb.exportKbZip('kb_abc123');
fs.writeFileSync('kb-export.zip', zip);

import { createWriteStream } from 'fs';

const stream = kb.exportKb('kb_abc123', 'jsonl'); // 'jsonl' (default) or 'csv'
const out = createWriteStream('kb-export.jsonl');
for await (const chunk of stream) {
  out.write(chunk);
}
out.end();