Veer Narayan@Veer78

PR Reviewer is a zero-friction AI code review tool built for the IBM Bob 2.0 Hackathon. THE PROBLEM Code reviews are slow, inconsistent, and error-prone — especially under time pressure. Developers reviewing their own PRs regularly miss security vulnerabilities, forget to write tests, and leave vague commit messages. Manual review doesn't scale. THE SOLUTION PR Reviewer lets you paste any raw git diff directly into a single web page — no GitHub login, no OAuth setup, no API tokens to configure. Hit submit, and Gemini 2.5 Flash analyses the diff and returns a structured three-section report: ⚠️ Risky / Breaking Changes — catches security flaws, signature bypasses, logic bugs, and breaking API changes 🧪 Missing Test Coverage — identifies untested code paths and new functions with no corresponding tests 📝 Unclear / Missing Commit Messages — flags vague, generic, or missing commit descriptions In our demo, the tool correctly caught a JWT decode signature bypass, a plaintext password comparison vulnerability, and a login function that always returned success — all explained clearly with no false positives. HOW IT WORKS The frontend is a single Next.js 15 App Router page with a textarea and a submit button. On submit, the diff is sent to a server-side API route (/api/review) which constructs a structured prompt and calls the Gemini 2.5 Flash API. The response is parsed into three labelled sections and rendered as bulleted result cards. The page includes a light/dark theme toggle with smooth gradient transitions. BUILT WITH IBM BOB 2.0 The entire project was designed and built using IBM Bob 2.0 as the AI coding assistant — from architecture planning in Plan mode (3 sub-tasks: scaffold, API route, UI) to full code generation in Agent mode. Bob autonomously resolved Next.js version vulnerabilities, handled gitignore conflicts during scaffolding, and iterated on the UI in real time. No code was written by hand. TECH STACK - Next.js 15 (App Router, TypeScript) - Tailwind CSS v3 - Google Gemini 2.5 Flash (@google/generative-ai) - Node.js server-side API route

Most trading-agent demos ask one model to analyze everything and place an order. That spends inference credits indiscriminately, hides disagreement, and lets generated text get too close to capital. TradeCouncil separates those responsibilities. Python first ranks a liquid watchlist using trend, moving averages, RSI, momentum, and volatility. Only the top one or two candidates reach the AI council, keeping the default cycle within six Featherless calls. A Bull Analyst builds the upside case, a Bear Analyst challenges it, and a CIO Judge returns a direction, confidence score, and rationale. Deterministic code then selects an eligible call or put and applies hard limits for DTE, delta, bid/ask spread, open interest, premium-at-risk, drawdown, and portfolio exposure. A model cannot override a Risk Sentinel veto. The full decision trail—quant ranking, adversarial debate, CIO verdict, contract, risk result, and execution receipt—is visible in one dashboard. The local competition workflow uses Alpaca market/trading APIs and the official Alpaca CLI against a dedicated paper account. The public judge walkthrough is intentionally credential-free, synthetic, and non-mutating, so anyone can inspect the experience without receiving broker authority. TradeCouncil's product path is a paper-first research and review workspace with saved risk policies, decision journals, broker reconciliation, and team analytics. It does not promise returns or passive income; options can expire worthless and involve substantial risk.

🌎 Hybrid Hackathon | 💻 Online Build: 12-17 October 2026 | On-site phase: 17-18 October 2026 To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. The AMD Developer Hackathon: ACT III is a hands-on event for developers, founders, engineers and builders who want to push what’s possible with AI on real infrastructure. 📍 If approved for the on-site experience, you'll also have the opportunity to showcase your project live. Travel and accommodation will not be covered.

💻 Online Hackathon | ⏳ Build: 48 hours | 📅 September 25–27, 2026 The official IBM Bob 2.0 hackathon. Build what's next in AI-assisted development, alongside the developers and experts shaping it. ⭐ Get help from experts throughout the event 🤝 Join a team or build solo 💰 Prize pool: $12,000 🧑🏻‍💻 Register before the Kick-Off Stream to start building with Bob

🌎 Online Hackathon | 💻 Dates: 28 August–4 September 2026 | 🏆 $6,000 Prize Pool Build AI trading agents on Alpaca — autonomous agents and trading apps using Alpaca's Trading API, MCP server and CLI. 🌍 Fully online hackathon. Join and build from anywhere in the world.