Detector should be configured on a Token contract which is has a peg to another token or a value (e.g. stablecoin, Liquid staking, LRT) Detector should monitor for token price deviation from an safe range. Detector should generate event for out of range condition.

🗂️ Bounty Category On-chain Security

💰 Prize Amount 1000

🥇 Number of Projects Awarded 1

🏆 Winner Breakdown 1st place: $1000

📝 Requirements Teams must develop a Depeg Monitor capable of detecting stablecoin and pegged asset deviations. The solution should:

Be configured on a token contract that maintains a peg (e.g., stablecoins, Liquid Staking Tokens (LSTs), Liquid Restaking Tokens (LRTs)). Define a safe range threshold for price variations and trigger an event when the token price falls outside this range. Use the Sentinel SDK (GitHub). ✅ What does a successful project look like? Detector successfully detects depeg

of an asset in a configured interval and generates an Alert with metadata describing depeg 🎨 UI/UX Design Requirements No front-end is required, the solution should be a command-line tool that runs autonomously. Must be packaged in Docker or as a standalone executable for Linux, MacOS, and Windows. 🧑‍⚖️ How are we judging it? How well does the detector identify depegging events? Are the alerts timely and reliable with minimal false positives? Is the code well-structured, efficient, and documented for easy maintenance and further development? How well is the Sentinel SDK integrated? 🌎 Impact on the organization Stablecoin depegging has caused big losses and financial instability in DeFi.A Depeg Monitor would provide real-time alerts, improving security and risk management for DeFi protocols, and liquidity providers. This tool will help protect on-chain assets by identifying risks before they escalate, making DeFi safer for all participants.

📚 Resources SDK: https://github.com/haas-labs/ext-sentinel-py-sdk Example: https://github.com/haas-labs/ext-sentinel-py-sdk/blob/main/examples/block_tx

👀 Some example use cases We are looking for proof-of-concept implementations that demonstrate a clear approach to detecting and reporting depegging events. For example: Track USDT, USDC, DAI, and other stablecoins against their pegged value to provide early warnings before major depegs occur.

💼 Recruitment Opportunities We are always looking for talented developers who are passionate about blockchain security. If your work on this bounty demonstrates strong technical skills and an understanding of real-time blockchain event detection, we encourage you to explore opportunities with our team.

�� Requirements The goal of this bounty is to develop an AI-powered security agent that detects vulnerabilities in Solidity smart contracts. The agent should be open-source, self-hostable, and capable of continuous learning using real-world security audit data. The MVP must feature a chat-based interface, allowing users to submit smart contracts via text input or GitHub integration for automated scanning. The AI model should be pre-trained on real audit reports and security vulnerabilities to ensure high accuracy.

✅ What does a successful project look like? A successful implementation of this bounty will result in a fully functional, AI-powered smart contract security auditor that can: ✅ Analyze Solidity smart contracts through a chat-based interface or GitHub repository integration ✅ Detect security vulnerabilities using static analysis + AI reasoning ✅ Provide actionable reports that classify risks and suggest mitigations ✅ Be self-hostable and extendable, allowing further training on private datasets

🎨 UI/UX Design Requirements The application must have a functional user interface, which can be:

A chat-based interface (web-app, CLI, or messaging bot). A browser-based dApp that allows users to submit smart contracts for analysis. For browser-based interfaces: Must be publicly accessible via a demo URL. Should include an input field for pasting Solidity code and an option to connect a GitHub repository. Results must be displayed clearly with vulnerability classifications and explanations. For command-line applications: Must be packaged in a Docker container for easy deployment. Should support Linux/MacOS environments.

🧑‍⚖️ How are we judging it? Success Criteria

1. AI-Powered Smart Contract Security Analysis (40%)

✅ Successfully accepts Solidity code and GitHub repositories as input. ✅ Identifies common vulnerabilities (reentrancy, overflows, access control issues, logic bugs). ✅ Uses a combination of static analysis tools (Slither, Semgrep, Foundry) + AI-based reasoning for enhanced accuracy. ✅ Provides explanations and context for each detected vulnerability. 2. AI Model Training & Security Knowledge (30%) ✅ Model is pre-trained on real-world audit reports provided in JSON/CSV format. ✅ Demonstrates understanding of security best practices by correctly classifying vulnerabilities. ✅ Supports additional training with new security datasets, allowing continued improvement. 3. Deployment & Usability (20%) ✅ Can be self-hosted on a VPS with a Docker-based deployment. ✅ Includes clear documentation on how to install, configure, and extend the system. ✅ Outputs reports in JSON, PDF, or Markdown format, making them easy to integrate into security workflows. 4. Innovation & Scalability (10%) ✅ Allows future customization by supporting additional datasets for fine-tuning. ✅ Demonstrates potential for integration into security pipelines for continuous auditing. ✅ Shows creative ways to enhance AI-powered vulnerability detection beyond static analysis. Example of a Successful Submission A team submits a project where:

Users can paste Solidity code or connect a GitHub repo via a chat-based interface. The AI agent scans the smart contract using static analysis + AI-powered reasoning to detect vulnerabilities. The system outputs a structured security report with severity classifications and suggested fixes. The pre-trained model leverages real-world audit data, improving its detection capabilities. The project is fully self-hostable on a VPS and can be further trained with additional datasets. Projects that meet or exceed these expectations will be eligible for the $7,000 (1st place) and $3,000 (2nd place) prizes.