Skip to content
RunStack

API, CLI and bot

An outside vantage point for scripts, CI and AI agents. Your agent runs inside your network — RunStack looks from outside, like your users and other mail servers do. The answer is a diagnosis, not raw data: verdict, findings, cause and fix.

API v1

dns, tls, http, email, port, ip — POST /api/v1/check/{tool}. The same checks as on the site, SSRF-safe and rate-limited.

Telegram bot

Send a domain to @RunStackru_bot — SSL, HTTP and email in one go.

CLI

A single binary for Linux, macOS and Windows: runstack check example.com. Exit codes for CI.

Quick start

  1. Get a personal key in your account — free, 200 checks a day.
  2. Call a check:
curl -s https://runstack.pro/api/v1/check/tls \
  -H "Authorization: Bearer $RUNSTACK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"host":"example.com"}'
{
  "ok": true,
  "tool": "tls",
  "verdict": "warn",
  "headline": "Certificate expires in 12 days",
  "findings": [{ "severity": "warning", "title": "…", "fix": "…" }],
  "reportUrl": "https://runstack.ru/tools/ssl-checker?host=example.com"
}

Full reference — API-01. Diagnosis language: ?lang=ru|en.

MCP for AI agents

Connect RunStack to Claude, Cursor and other agents as an MCP server (Streamable HTTP). Tools: check_tls, check_http, check_email_domain, check_dns, check_port, ip_lookup. The agent gets a diagnosis with cause, fix and a how-to link.

{
  "mcpServers": {
    "runstack": {
      "type": "http",
      "url": "https://runstack.pro/api/mcp",
      "headers": { "Authorization": "Bearer rsk_..." }
    }
  }
}

Same key as the API (your account). Without a key the server lists tools but does not run checks.

Need more?

We're preparing higher limits and a GitHub Actions / GitLab CI example. Tell us how you'd use it — it decides what we build first.

Use case