AI-Powered Email Verification
Connect filtrat.io to your AI assistant. Ask it to verify, find, or check emails in plain English. It handles the rest.
Works with Claude, Cursor, Windsurf, and any MCP-compatible client.Connect in one step
Add filtrat to your AI client's MCP config file. Claude Desktop connects through the mcp-remote helper, which runs with npx (Node.js 18 or newer). Cursor, Windsurf and other clients that take a server URL use the second config.
- Replace
YOUR_API_KEYwith yourak_key → restart your client → done. - Or just copy the config and ask your AI agent to add it to your config file.
- Get your key: API page → Keys (the MCP tab has these configs with your key filled in)
{
"mcpServers": {
"filtrat": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.filtrat.io/mcp",
"--header",
"Authorization: Bearer YOUR_API_KEY"
]
}
}
}{
"mcpServers": {
"filtrat": {
"url": "https://mcp.filtrat.io/mcp",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}Just ask.
filtrat does the work.
Paid actions preview first: the agent shows you what will run and its credits. Anything over 5,000 credits starts only after you confirm.
What's available
16 tools, one way to do each thing.
Lists are addressed by name, every reply tells the agent what to call next, and capture saves whole searches — no 500-lead limit.
verify
One address or many. Deliverable, invalid or catch-all — resolved. Up to 25 come back at once; more run as a list you follow.
per verified or invalid result; +5 when a catch-all is resolved
find
Name + company or domain → verified work email, for one person, many, or a whole people list. Not found — free.
per email found (25 with catch-all resolution)
search_linkedin
Explore LinkedIn: people, companies, posts, events, jobs or groups. Returns the count and samples; never saves.
a company filter given by LinkedIn URL: 20 per company looked up (free again for 30 days)
lookup_linkedin
Read one profile, company page, post or event.
capture
Save the people from a Sales Navigator search, a post, an event, a saved search, or the decision-makers at a company list. Previews first; no 500 limit — searches over 2,500 split automatically.
searches and attendees free; post engagers 10 per profile; at a company list, 20 per company added only by its LinkedIn URL (looked up once)
get_list
A list by name: 25 rows at a time, what's still running on it, credits used, or CSV.
manage_list
Create, rename, delete, add rows (your own companies or people, up to 1,000 a call), remove rows.
manage_job
Pause, resume, cancel or retry what runs on a list.
enrich
Match imported companies to LinkedIn, fill company data or profiles, or answer an AI question for every row.
company data per company filled; matching 20–150 per company found (50 by domain); AI question 10 per value (Growth plan and up)
targeting
Your company profile and ICPs: show, set up from text or your website, edit. Growth plan and up.
score
Score a list against an ICP, a sample first if you like. Growth plan and up.
per row
watch
Keep watching a profile, company page, post, event, group or keywords. Pro plan and up.
post engagers 10 per profile
watchers
List, read (signals too), pause, resume, stop or check your watchers.
comment_on_post
Comment from your own LinkedIn account. Every comment shows you the exact text first and posts only after you approve it.
write_outreach
Three personalised openers for a lead. Growth plan and up.
get_credits
Your balance and what recent work cost.
Ready-made starting points
Eight prompts in your agent's prompt menu.
leads_from_linkedin
leads_from_post_or_event
find_emails_for_a_list
verify_a_list
watch_linkedin
score_my_list
decision_makers_at_my_companies
set_up_targeting
See it in action
A 10,000-email file in, a clean list out. The agent submits the bulk job, follows it and reports what it removed.
Reading campaign.csv... 10,000 emails found.
Submitting bulk verification job...
→ Job started: job_8f2a4c91
→ Processing: 10,000 emails
Checking progress...
→ 10,000 / 10,000 complete
Results:
→ 7,842 deliverable
→ 1,456 invalid (will bounce)
→ 512 catch-all (uncertain)
→ 190 unknown
✓ Done. 9,298 credits used.
Removed 1,456 bad addresses.
Your list is ready to send.
Same credits.
No MCP markup.
Same balance across dashboard, API, and MCP. Plans from $10/mo. 200 free credits to start.
Full API reference: filtrat.io/v1/docs
200 free credits.
No card.
Copy the config, replace your key, and start verifying in seconds.