DEV Community

Fred Santos
Fred Santos

Posted on

5 Utility Endpoints Every AI Agent Needs: From Sentiment to Summarization

#webdev #ai #productivity #api

AI agents are incredibly powerful — but they're often bottlenecked by the same five utility tasks: understanding text sentiment, summarizing long documents, looking up domain info, checking public holidays, and querying DNS records.

These are the "plumbing" tasks. Boring? Maybe. But without them, your agent can't answer half the real-world questions users throw at it.

This post covers five endpoints we recently shipped on IteraTools that handle these tasks — all pay-per-use, no setup, available to any HTTP client.

1. POST /sentiment — Understand the emotional tone of text

Before your agent acts on user feedback, it should know how the user feels.

curl -X POST https://api.iteratools.com/sentiment \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "I love this product, it works great!"}'
Enter fullscreen mode Exit fullscreen mode

Response:

{
  "sentiment": "positive",
  "score": 7,
  "comparative": 1.4,
  "positive": ["love", "great"],
  "negative": [],
  "language": "en"
}
Enter fullscreen mode Exit fullscreen mode

Uses AFINN-165 — runs entirely offline, no third-party NLP API. Perfect for analyzing customer reviews, support tickets, or social media comments at scale.

Price: $0.001 per call | Docs: iteratools.com/tools/sentiment

2. POST /summarize — Extract key sentences from any text or URL

Long documents are everywhere. Your agent shouldn't have to read a 5,000-word article just to answer a simple question.

# Summarize a URL
curl -X POST https://api.iteratools.com/summarize \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://en.wikipedia.org/wiki/Machine_learning", "sentences": 3}'
Enter fullscreen mode Exit fullscreen mode 
# Or summarize raw text
curl -X POST https://api.iteratools.com/summarize \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Artificial intelligence is transforming the world. Machine learning enables computers to learn from data. Deep learning uses neural networks with many layers. Natural language processing helps computers understand human language. These technologies are changing how we work and live.",
    "sentences": 2
  }'
Enter fullscreen mode Exit fullscreen mode

Response:

{
  "summary": "Artificial intelligence is transforming the world. Natural language processing helps computers understand human language.",
  "sentences": 2,
  "original_length": 283,
  "compression_ratio": 0.4276
}
Enter fullscreen mode Exit fullscreen mode

This uses extractive summarization — it picks real sentences from the original text, scored by TF-IDF frequency + position weighting. No hallucinations, no rewrites. Supports up to 50,000 characters input.

Price: $0.002 per call | Docs: iteratools.com/tools/summarize

3. POST /dns/lookup — Query DNS records for any domain

Need to verify an MX record before sending email? Check if a site has DNSSEC? Audit nameservers?

curl -X POST https://api.iteratools.com/dns/lookup \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain": "github.com", "types": ["A", "MX", "TXT"]}'
Enter fullscreen mode Exit fullscreen mode

Response:

{
  "domain": "github.com",
  "records": {
    "A": ["140.82.114.4"],
    "MX": [{"exchange": "aspmx.l.google.com", "priority": 1}],
    "TXT": ["v=spf1 ip4:192.30.252.0/22 include:_netblocks.google.com ~all"]
  }
}
Enter fullscreen mode Exit fullscreen mode

Supports A, AAAA, MX, TXT, CNAME, NS, SOA, and PTR records — up to 8 types per request.

Price: $0.001 per call | Docs: iteratools.com/tools/dns-lookup

4. GET /calendar/holidays — Public holidays for 100+ countries

Building scheduling logic? Automating invoices? Your agent needs to know when the world is closed.

curl "https://api.iteratools.com/calendar/holidays?countryCode=BR&year=2026" \
  -H "Authorization: Bearer YOUR_KEY"
Enter fullscreen mode Exit fullscreen mode

Response:

{
  "year": 2026,
  "countryCode": "BR",
  "holidays": [
    {"date": "2026-01-01", "name": "New Year's Day", "localName": "Ano novo"},
    {"date": "2026-04-21", "name": "Tiradentes", "localName": "Tiradentes"},
    ...
  ],
  "count": 12
}
Enter fullscreen mode Exit fullscreen mode

Powered by the Nager.Date API. Supports 100+ country codes with ISO 3166-1 alpha-2 format.

Price: $0.001 per call | Docs: iteratools.com/tools/calendar-holidays

5. POST /whois — Domain ownership and expiration info

Is a domain available? When does it expire? Who's the registrar? Your agent can answer all of this.

curl -X POST https://api.iteratools.com/whois \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain": "iteratools.com"}'
Enter fullscreen mode Exit fullscreen mode

Response:

{
  "domain": "iteratools.com",
  "registrar": "Namecheap, Inc.",
  "created": "2024-11-15",
  "expires": "2025-11-15",
  "status": ["clientTransferProhibited"],
  "nameservers": ["dns1.registrar-servers.com", "dns2.registrar-servers.com"],
  "available": false
}
Enter fullscreen mode Exit fullscreen mode

Uses RDAP (the modern WHOIS replacement) with a structured JSON fallback. No raw WHOIS text to parse.

Price: $0.001 per call | Docs: iteratools.com/tools/whois

Putting it together: A Python agent that researches a domain

Here's how you'd combine all five in a single Python agent workflow:

import httpx

API_KEY = "YOUR_ITERATOOLS_KEY"
BASE = "https://api.iteratools.com"
HEADERS = {"Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json"}

def research_domain(domain: str, url: str) -> dict:
    results = {}

    # 1. Summarize the website content
    r = httpx.post(f"{BASE}/summarize", headers=HEADERS, json={"url": url, "sentences": 3})
    results["summary"] = r.json().get("summary", "")

    # 2. Analyze sentiment of the summary
    r = httpx.post(f"{BASE}/sentiment", headers=HEADERS, json={"text": results["summary"]})
    results["sentiment"] = r.json().get("sentiment", "neutral")

    # 3. WHOIS lookup
    r = httpx.post(f"{BASE}/whois", headers=HEADERS, json={"domain": domain})
    whois = r.json()
    results["registrar"] = whois.get("registrar")
    results["expires"] = whois.get("expires")

    # 4. DNS records
    r = httpx.post(f"{BASE}/dns/lookup", headers=HEADERS, json={"domain": domain, "types": ["A", "MX"]})
    results["dns"] = r.json().get("records", {})

    # 5. Check holidays for today's country (if we knew it)
    r = httpx.get(f"{BASE}/calendar/holidays?countryCode=US&year=2026", headers=HEADERS)
    results["us_holidays_2026"] = r.json().get("count", 0)

    return results

report = research_domain("iteratools.com", "https://iteratools.com")
print(report)
Enter fullscreen mode Exit fullscreen mode

Total cost for this workflow: ~$0.007 (less than a cent).

One key, 36 tools

All of these endpoints — and 31 more — are available under a single API key at iteratools.com. No subscriptions, no free-tier limits, no juggling multiple providers.

Pay only what you use. Start with $1 and get hundreds of API calls.

→ Get your key | Browse all 36 tools | API Docs

Top comments (0)