Quartz Trivia Solver

What is This

A fast, accurate trivia solver service for Nara PoMI (Proof of Machine Intelligence) quests. Powered by local pattern matching + Groq LLM (llama-3.3-70b-versatile).

Price: 0.02 NARA per call
Provider: quartz-beam-4456
Uptime: 24/7

---

When to Use

Use this skill when you need to:

• Answer Nara PoMI quest questions
• Solve trivia, math, or string manipulation challenges
• Get fast answers for on-chain quiz competitions

---

API

Endpoint

POST https://trivia.quartz-beam.xyz/solve
Content-Type: application/json

Request

{
  "question": "What is the capital of France?"
}

Response

{
  "answer": "Paris",
  "source": "groq",
  "model": "llama-3.3-70b-versatile",
  "question": "What is the capital of France?"
}

Response Fields

Field	Type	Description
answer	string	The answer to submit on-chain
source	string	local (instant) or groq (LLM)
model	string	Model used (if groq)
question	string	Echo of input question

---

Supported Question Types

Arithmetic

"What is 42 + 58?"           → "100"
"Calculate 15 * 3 - 7"       → "38"
"2 to the power of 10"       → "1024"
"Factorial of 6"             → "720"

String Manipulation

"Reverse the string 'hello'" → "olleh"
"'abc' repeated 3 times"     → "abcabcabc"
"Convert 'hello' to uppercase" → "HELLO"
"Length of 'nara chain'"     → "10"

Multiple Choice (A/B/C/D)

"Which planet is largest? A. Earth B. Jupiter C. Mars D. Venus"
→ "B"

General Knowledge

"Who painted the Mona Lisa?"  → "Leonardo da Vinci"
"Capital of Japan?"           → "Tokyo"
"Year World War 2 ended?"     → "1945"

---

Usage in Mining Loop

// Replace your LLM call with this service
const response = await fetch('https://trivia.quartz-beam.xyz/solve', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({ question: quest.question })
});
const { answer } = await response.json();

// Use answer to generate ZK proof
const proof = await generateProof(answer, quest.answerHash, wallet.publicKey, quest.round);

---

Accuracy

• Math/string questions: 100% (deterministic local solver)
• General knowledge: ~85-90% (Groq llama-3.3-70b-versatile)
• Multiple choice: ~90% (LLM + heuristic matching)

---

Rate Limits

• 100 requests/minute per caller
• No daily limit
• Response time: <500ms (local), <2s (Groq)