devdrops-prediction-markets— agent skill

When to use this skill

• User wants to know the current market probability for a real-world event
• Agent is building a decision-support tool, news tracker, or probability aggregator
• User asks "what does the market say about [event]?", "probability of [outcome]?", or "find markets about [topic]"
• Agent needs calibrated crowd forecasts to augment its own analysis

How it works

The endpoint aggregates live markets from Polymarket (on-chain prediction markets) and Manifold Markets (play-money forecasting platform). Markets are normalised to a common schema: each has a title, probability (0–1), platform, close date, and volume. The /markets endpoint lists active markets sorted by volume; /search filters by keyword. This endpoint costs $0.005 per query.

Endpoints

Parameters

Response fields

Quick start

TypeScript

import { wrapFetchWithPayment } from "@x402/fetch";
const pay = wrapFetchWithPayment(fetch, walletClient);

// Top markets by volume
const res = await pay("https://api.devdrops.run/api/predictions/markets?limit=10");
const { data } = await res.json();

// Search for election markets
const searchRes = await pay(
  "https://api.devdrops.run/api/predictions/search?q=election+2026"
);

curl

curl -H "x-402-payment: <proof>" \
  "https://api.devdrops.run/api/predictions/markets?limit=5"

Examples

Example 1: Top markets

GET /api/predictions/markets?limit=3

{
  "product": "predictions",
  "data": {
    "markets": [
      {
        "id": "poly-0x1a2b",
        "platform": "polymarket",
        "title": "Will the Fed cut rates at the June 2026 meeting?",
        "probability": 0.34,
        "volume": 2840000,
        "close_date": "2026-06-12",
        "url": "https://polymarket.com/event/fed-june-2026"
      },
      {
        "id": "mf-abc123",
        "platform": "manifold",
        "title": "Will OpenAI release GPT-5 before July 2026?",
        "probability": 0.61,
        "volume": 45000,
        "close_date": "2026-07-01",
        "url": "https://manifold.markets/question/openai-gpt5-2026"
      }
    ]
  }
}

Example 2: Search for a specific topic

GET /api/predictions/search?q=bitcoin+price+2026&platform=polymarket

Returns Polymarket markets matching "bitcoin price 2026".

Example 3: Get a specific market

GET /api/predictions/market/poly-0x1a2b

Returns full details for a single market including order book if available.

Example 4: Decision support agent

async function getConsensus(topic: string): Promise<string> {
  const res = await pay(
    `https://api.devdrops.run/api/predictions/search?q=${encodeURIComponent(topic)}&limit=5`
  );
  const { data } = await res.json();

  if (!data.markets.length) return `No prediction markets found for: ${topic}`;

  const avg = data.markets.reduce((s: number, m: any) => s + m.probability, 0) / data.markets.length;
  const topMarket = data.markets.sort((a: any, b: any) => b.volume - a.volume)[0];

  return `Market consensus on "${topic}": ${(avg * 100).toFixed(0)}% average probability. ` +
    `Highest-volume market: "${topMarket.title}" at ${(topMarket.probability * 100).toFixed(0)}%.`;
}

Example 5: Cross-platform probability comparison

const [poly, manifold] = await Promise.all([
  pay(`https://api.devdrops.run/api/predictions/search?q=AI+regulation&platform=polymarket`),
  pay(`https://api.devdrops.run/api/predictions/search?q=AI+regulation&platform=manifold`),
]);

const polyProbs = (await poly.json()).data.markets.map((m: any) => m.probability);
const manifoldProbs = (await manifold.json()).data.markets.map((m: any) => m.probability);

Errors & limits

No free tier. All requests $0.005.

Caching: Market data cached 5 minutes (shorter than other products due to live pricing).

Probability scale: All probabilities normalised to 0–1 regardless of platform. Manifold uses a Bayesian scoring rule; Polymarket uses LMSR or orderbook pricing. Direct comparison is approximate.

Resolved markets: Resolved markets may still appear briefly with probability 0 or 1. Filter on close_date > today to see only active markets.

Related skills

• devdrops-fx-rates — pair with prediction markets for macro context
• devdrops-research-brief — synthesise market data into a structured brief