.png&w=256&q=75)
Vishwa Kumaresh@theversion2
theversion2
🤓 Latest Submissions
ProofRank
ProofRank is an AI diligence engine for hackathon judges, sponsor teams, accelerators, and grant reviewers who need to verify many public AI projects under time pressure. It reviews project pages, GitHub repositories, demos, and technology claims, then ranks submissions with source-backed proof receipts. The core workflow is built around Bright Data as the evidence acquisition layer. ProofRank’s verified sponsor receipt uses Bright Data Remote MCP tools including scrape_as_markdown, search_engine, and discover to collect source, search, and discovery evidence. Each proof receipt records trace state, provider, byte count, content hash, trace digest, and signature status so judges can distinguish executed evidence from claimed or planned integrations. The app includes a ranked queue, Evidence Route, Claim Ledger, Proof Receipt panel, Originality Radar, adversarial review notes, exportable packets, and a secure live-review API fallback. It was built and published through native.builder, with Vercel used as a fallback live app and API shell. ProofRank helps judges move from subjective browsing to repeatable, evidence-backed review.
10 Aug 2026
Ridhwan
AI agents are becoming autonomous economic actors — trading tokens, deploying contracts, managing treasuries. But enterprises won't trust them with real money without guardrails. There's no governance layer ensuring these agents follow policies, stay within budgets, or produce auditable compliance trails. Ridhwan solves this. It's an enterprise trust and commerce mesh that sits between your AI agent and the blockchain, enforcing policies, tracking budgets, auditing every action, and managing risk — all in real time. Built on SURGE and Base (Coinbase L2), Ridhwan provides: - A universal policy engine with budget caps, risk scoring, and human-in-the-loop approvals - SURGE wallet integration for on-chain token launches, transfers, and trading - x402 gasless transactions so agents can transact without managing gas - An immutable audit ledger with full compliance receipts - GNN-based fraud detection using graph neural network message-passing - A multi-agent orchestrator coordinating 5 specialized sub-agents for risk, policy, trading, compliance, and trust - x402 autonomous commerce enabling agent-to-agent paid services via the HTTP 402 protocol - Trust delegation with constraint-scoped capability chains and cascade revocation - Decentralized identity with Ed25519 DIDs and zero-knowledge proofs - Smart contract verification with vulnerability scanning - A web dashboard with 15 sections and demo mode for presentations The system exposes 130+ REST endpoints, supports real-time SSE event streaming, and integrates with Moltbook for automated build updates. All API keys are optional — Ridhwan runs in full simulation mode without them. Built with TypeScript, OpenClaw, SURGE on Base, and the x402 Protocol.
28 Feb 2026
