Prism AI is a conversational AI assistant built on top of DomainScan's full diagnostic catalog. Ask in plain English — about a domain, an IP address, an SPF record, a JWT token, a TLS certificate, an email deliverability problem — and Prism AI selects the right tools from a library of 50+ real diagnostic functions, runs them in parallel, and synthesizes the results into a complete answer. Every tool invocation is visible in the right panel: which tool ran, how long it took, and exactly what it returned. General AI assistants can discuss network concepts; Prism AI fetches the actual data and reasons over it.
A general AI assistant tells you what an SPF record is. Prism AI fetches yours, parses it, expands every include, counts your DNS lookups against the limit, cross-references your DMARC policy, checks whether your sending IPs are blacklisted, and tells you exactly why your mail is bouncing. Six things that make this possible:
Prism AI doesn't hallucinate WHOIS records or imagine what your SSL certificate looks like. It runs the same diagnostic tools you'd run manually — but selects them automatically based on your question. The data is real, current, and fetched fresh.
Every public tool on DomainScan is wired into Prism AI: WHOIS, DNS, SSL, security headers, SPF/DKIM/DMARC, blacklist, port scanning, traceroute, JWT decoding, hash generation, and more. The right tool selection happens automatically.
Complex questions get multi-tool answers. "Analyze my website security" triggers SSL Info, Security Headers, DMARC, DNS, and IP Blacklist in parallel — 5 tools, ~22 seconds — then synthesizes one coherent answer instead of five disconnected reports.
Every tool call is visible in real time: which tool ran, what it took, exactly what it returned. Click 'View Full' on any result to see the raw JSON. No black box — you can audit how Prism AI reached its conclusions.
Ask a follow-up — "so what should I change first?" or "compare to cloudflare.com" — and Prism AI extends the analysis, calling more tools as needed and preserving context across turns. The conversation builds; it doesn't restart with every question.
Below the input: Analyze domain security, Check SSL, DNS lookup, Port scan, WHOIS, JWT decoder, and more. For common questions you don't even need to type — pick the pattern and Prism AI handles the rest.
Ask ChatGPT, Claude, or Gemini "why is my email going to spam?" — you'll get a thoughtful, well-written, mostly-correct overview of common causes (SPF, DKIM, DMARC, sender reputation, content). What you won't get: an actual check of YOUR SPF record, YOUR DMARC policy, or YOUR IP's blacklist status. The general AI doesn't have access to those tools. It can only tell you what to check; it can't check.
Behind every Prism AI answer is a structured pipeline. You see the question and the answer; the right panel shows the middle steps. Here's what happens between them:
Prism AI handles every diagnostic question DomainScan's 50+ tools can answer — and many that require multiple tools working together. Twelve recurring question types, with sample prompts you can paste directly:
"Analyze the security of stripe.com." → SSL/TLS posture, security headers (HSTS, CSP, X-Frame-Options), DMARC policy, IP blacklist status, DNS health. Returns a prioritized list of findings.
"Why are my emails going to spam? My domain is acme.com." → SPF (lookup count, qualifier), DKIM (selectors, key length), DMARC (policy, alignment), reverse DNS, IP blacklist. Returns the specific failure points.
"What SSL certificate does github.com use?" → Issuer, validity dates, subject alternative names, key algorithm, supported TLS versions, weak-protocol warnings. Surfaces deprecated TLS 1.0/1.1 if still enabled.
"Tell me everything about example.com." → WHOIS/RDAP, age, registrar, nameservers, DNS records, hosting infrastructure, SSL setup, redirect chains. A complete fingerprint in one pass.
"Compare the security setup of cloudflare.com and google.com." → Side-by-side: SSL strength, header coverage, email auth, blacklist status. Identifies where each domain is stronger or weaker.
"What's hosted at 172.66.0.227?" → Reverse DNS, hosted domains, blacklist status across 50+ databases, geolocation, ASN, ping latency. Useful for incident response and abuse investigation.
"Why isn't www.example.com resolving?" → Full DNS record dump (A, AAAA, CNAME, MX, NS, TXT, SOA), authority chain, nameserver consistency, DNSSEC status, propagation check.
"Is my site discoverable by ChatGPT and Perplexity?" → robots.txt for AI crawlers (GPTBot, ClaudeBot, PerplexityBot), llms.txt presence, server-side rendering check, structured data audit. Tells you what AI search will and won't see.
"Decode this JWT for me: eyJhbGciOi..." → Header, payload, claims, expiration status, signature info. Combined with: "is it expired?", "what's the audience?", "what algorithm signed it?"
"How fast is google.com from different regions?" → Multi-region ping, latency stats, packet loss, jitter. Useful for diagnosing region-specific connectivity issues.
"I received a phishing email from [email protected]. Investigate." → WHOIS age (recently registered = red flag), MX records, SPF/DMARC posture, hosting infrastructure, similarity to legitimate brands. Surfaces the indicators of compromise.
"Run a full security and infrastructure report on acquisition-target.com." → Multi-section deep dive covering email auth, TLS, headers, DNS, hosting, registrar history. The kind of report that takes a security analyst an hour, delivered in 30 seconds.
Prism AI can invoke any of DomainScan's 50+ tools. Every tool is also available as a standalone page if you want to run it directly. Six categories:
Four concrete examples — the kind of multi-turn conversation that demonstrates what Prism AI does that simpler tools can't:
Prism AI doesn't replace general AI assistants or specialized network tools — it fills a gap between them. Honest comparison of where each one belongs:
Most AI agents are black boxes. You ask a question, the agent works on it, you get an answer — and you have no idea what data the answer is based on, whether the right tools ran, or how to verify any specific claim. Prism AI is built the opposite way. Every action is visible in real time:
Prism AI works on any plain-English question, but a few patterns consistently produce better answers. Six things that move quality up:
Prism AI's conversational interface is available as a streaming JSON API for embedding in your own tools, runbooks, security automations, or chat platforms. The agent handles tool selection, parallel execution, and synthesis on its end; your code sends a question and receives the answer plus the trail of tool invocations.
// Single-turn query
const res = await fetch('https://api.domainscan.in/v1/prism/chat', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
message: 'Analyze the security of stripe.com',
stream: false
})
});
const data = await res.json();
console.log(data.answer); // synthesized text answer
console.log(data.toolsUsed); // [{ name: 'ssl_info_lookup', durationMs: 117 }, ...]
console.log(data.toolResults); // raw outputs from each tool, keyed by tool name
// Multi-turn conversation — pass a conversationId to preserve context
const res1 = await fetch('https://api.domainscan.in/v1/prism/chat', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
message: 'Analyze cloudflare.com security'
})
}).then(r => r.json());
const res2 = await fetch('https://api.domainscan.in/v1/prism/chat', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
message: 'Now compare it with akamai.com',
conversationId: res1.conversationId // reuse context
})
}).then(r => r.json());
console.log(res2.answer); // comparison reuses cloudflare data from turn 1
// Streaming response — useful for chat UIs
const stream = await fetch('https://api.domainscan.in/v1/prism/chat', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
message: 'Why is my email going to spam? Domain: acme.com',
stream: true
})
});
const reader = stream.body.getReader();
const decoder = new TextDecoder();
while (true) {
const { value, done } = await reader.read();
if (done) break;
const events = decoder.decode(value).split('\n\n').filter(Boolean);
for (const event of events) {
const data = JSON.parse(event.replace(/^data: /, ''));
// events: { type: 'tool_start', tool: 'spf_check' }
// events: { type: 'tool_complete', tool: 'spf_check', durationMs: 234 }
// events: { type: 'token', delta: 'Your SPF record has...' }
// events: { type: 'done', toolsUsed: [...] }
console.log(data);
}
}{
"success": "boolean",
"conversationId": "string (UUID — pass back to preserve context)",
"answer": "string (Prism AI's synthesized response)",
"toolsUsed": [
{
"name": "string (e.g. 'ssl_info_lookup', 'spf_check')",
"description": "string (human-readable description)",
"durationMs": "number",
"status": "completed | failed",
"error": "string | null"
}
],
"toolResults": {
"<tool_name>": "object (the raw JSON output from each tool)"
},
"totalDurationMs": "number (wall-clock time for the full response)"
}
// Request body
{
"message": "string (required — the user's question)",
"conversationId": "string (optional — pass to continue a previous conversation)",
"stream": "boolean (optional — defaults to false)"
}