CircuitX

Core Problems

1. Strategy Exposure & Front-Running
2. Lack of Privacy in DeFi Trading
3. Vulnerability to Market Manipulation

Our solution

• Complete privacy through ZK proofs. Traders prove position validity without revealing any details. Trade with confidence knowing strategies stay private.
• Position details are cryptographically hidden. Only commitment hashes are on-chain—impossible to reverse-engineer. Traders can execute large positions without revealing strategy or becoming targets.
• Private positions cannot be front-run or targeted. MEV bots cannot see trades, eliminating sandwich attacks and front-running

What People Can Use CircuitX For

Retail Traders

• Strategy Protection: Develop and execute trading strategies without exposure
• Privacy-Conscious Trading: Trade without revealing financial position
• MEV Protection: Avoid front-running and sandwich attacks
• Competitive Advantage: Keep successful strategies private
• Example: A retail trader discovers a profitable arbitrage opportunity. On CircuitX, they can execute it repeatedly without competitors seeing the pattern and copying it.

How CircuitX Makes Existing Tasks Easier/Safer

• No Strategy Management Overhead
• Simplified Risk Management
• Reduced Monitoring
• Strategy Security

Challenges I Ran Into

1. Transaction Nonce Mismatch Errors

• Problem: When calling lock_collateral followed immediately by open_position, the second transaction failed with “Invalid transaction nonce” because the account nonce hadn’t updated on-chain yet

• Solution: Implemented nonce polling—after lock_collateral confirms, poll the account’s nonce every 500ms until it increments before sending the second transaction. Added a 2-second fallback delay as a safety measure.

2. Commitment Format Mismatch (COMMITMENT_MISMATCH Error)

• Problem: When closing positions, the commitment in the proof’s public inputs didn’t match the commitment parameter passed separately, causing COMMITMENT_MISMATCH errors.

• Solution: Changed the flow to generate the proof first, then extract the commitment directly from proofResult.publicInputs[1] instead of normalizing the stored position commitment separately. This ensures both commitments match exactly.

3. Market ID Format Inconsistencies

• Problem: Different parts of the system used different formats for market IDs (string vs felt252), causing validation failures.

• Solution: Created a shared getMarketIdFelt() function to ensure consistent format across the codebase, and removed strict market ID validation since the contract handles format variations internally.