Introduction

Blockchains are Deterministic. The Real World is Non-Deterministic. Therefore, a Blockchain cannot natively know the price of Tesla stock, the temperature in London, or the result of the Super Bowl. If it did, nodes would disagree, and Consensus would break.

This lesson explores the mechanism of Oracles: The secure bridge between the Chaos of Reality and the Order of the Ledger.

---

The Physics: The Oracle Problem

Why can’t smart_contract.get("http://google.com") work? Because Node A might get “200 OK” and Node B in China might get “400 Bad Request”. The chain would fork.

The Solution: Data must be pushed into the block by a transaction. Once it is in the block, it is immutable history. Every node sees the same data. The Oracle is not the data source; it is the entity that signs the transaction containing the data.

---

Deep Dive: Chainlink (Decentralized Aggregation)

A single reporting node is a Single Point of Failure (SPOF). Chainlink solves this with a Decentralized Network of Nodes (DON).

The Physics:

1. Observation: 31 Nodes fetch the price of ETH from Binance, Coinbase, Kraken.
2. Reporting: They share values off-chain (P2P).
3. Aggregation: They calculate the Median? (Resistant to outliers/corruption).
4. Consensus: They generate a cryptographic proof (BLS Signature).
5. Transmission: One node submits the aggregated value + signature to the blockchain.

Result: You trust the Network Consensus, not a single node.

---

Strategy: TWAP (Uniswap Time-Weighted Average Price)

Uniswap pools contain price data. Can we use them as Oracles? Yes, but not the Spot Price.

Flash Loan Attack:

1. Attacker borrows $100M.
2. Buys ETH on Uniswap (Price spikes to $5000).
3. Your contract reads Spot Price ($5000).
4. Attacker borrows against inflated collateral.
5. Attacker sells ETH back (Price returns to $2000).

Defense (TWAP): The price is averaged over the last 30 minutes. To manipulate the TWAP, the attacker must hold the price high for 30 minutes, costing millions in arbitrage losses.

---

Code: Reading a Chainlink Feed

import "@chainlink/contracts/src/v0.8/interfaces/AggregatorV3Interface.sol";

contract PriceConsumer {
    AggregatorV3Interface internal priceFeed;

    constructor() {
        // ETH/USD Feed on Mainnet
        priceFeed = AggregatorV3Interface(0x5f4eC3Df9cbd43714FE2740f5E3616155c5b8419);
    }

    function getLatestPrice() public view returns (int) {
        (
            uint80 roundID, 
            int price,
            uint startedAt,
            uint timeStamp,
            uint80 answeredInRound
        ) = priceFeed.latestRoundData();
        
        // Sanity Check: Is data stale?
        require(timeStamp > block.timestamp - 3600, "Stale Price");
        require(price > 0, "Negative Price");
        
        return price;
    }
}

---

Practice Exercises

Exercise 1: The Outlier (Beginner)

Scenario: 5 Nodes report: [100, 101, 100, 9999, 100]. Task: Calculate Median. (Answer: 100). Calculate Average. (Answer: 2080). Lesson: Why Oracles use Medians, not Averages.

Exercise 2: Cost of Corruption (Intermediate)

Scenario: A protocol secures $1B TVL effectively. **Oracle:** Controlled by 3 nodes. Bribing 2 nodes costs$1M. Result: $1B >$1M. The Oracle WILL be corrupted. Security requires $Cost(Corruption) > Profit(Attack)$.

Exercise 3: Staleness Check (Advanced)

Scenario: L2 Network goes down. Oracle stops updating. Task: Your contract reads the price. It gets the value from 5 hours ago (Pre-crash). Defense: Always check block.timestamp - lastUpdateTimestamp < Threshold.

---

Knowledge Check

1. Why can’t blockchains make HTTP requests?
2. What is the difference between Spot Price and TWAP?
3. Why does Chainlink use the Median?
4. What prevents a Chainlink node from lying?
5. What is “Heartbeat” in an oracle feed?

Answers

1. Non-Determinism. External world state changes, breaking consensus.
2. Time. Spot is instant (easily manipulated). TWAP is historical average (expensive to manipulate).
3. Robustness. Median filters out extreme outliers/malicious reports.
4. Staking/Reputation. Provide bad data -> Lose money/jobs.
5. Frequency. The guaranteed update time (e.g., “Every 1 hour” or “Every 0.5% deviation”).

---

Summary

• Oracle: Truth API.
• Median: Corruption Filter.
• TWAP: Flash Loan Shield.

---