NEXUS AI trading agent

NEXUS is a trust-aware autonomous AI trading agent built for the next generation of financial agents that must do more than just trade. It was designed around a core challenge in AI finance: enabling agents to interact with capital safely, execute strategies autonomously, and demonstrate transparent, verifiable behavior. The project combines three strategy modes (algo, llm, hybrid), four risk profiles, DEX-based execution through a RiskRouter-compatible flow, ERC-8004 agent identity, and EIP-712 checkpoint logging. This means NEXUS is not just an automated trader, but a system built to make every important decision inspectable and accountable. One of its biggest strengths is trust infrastructure. NEXUS aligns with the ERC-8004 vision of identity, reputation, and validation for financial agents. Each agent can operate with a registered on-chain identity, while decision checkpoints create an auditable trail of actions, confidence, and reasoning. That gives the system a verifiable record instead of a black-box trading loop. Another major strength is risk-aware execution. NEXUS is built around policy controls rather than blind automation. It includes risk profiles, decision guards, a local kill-switch, recovery controls, and checkpointed runtime behavior so the agent can prioritize capital protection, drawdown awareness, and safer execution. This matches the hackathon’s emphasis on risk-adjusted performance, validation quality, and transparent agent behavior. NEXUS also stands out through its modular architecture. It combines PRISM-backed market and signal inputs, Groq-hosted Llama reasoning, algorithmic scoring modules, and a live Next.js dashboard for monitoring logs, checkpoints, system state, and controls in real time. In short, NEXUS shows that an autonomous trading agent can be configurable, transparent, risk-aware, and verifiable by design.

Aegis

Clawgov

ClawGov is an end-to-end governance toolkit that brings decentralized security and community oversight to OpenClaw AI agents. Built for the SURGE × OpenClaw Hackathon, it acts as an intelligent firewall and governance layer that prevents agents from executing dangerous or unapproved community skills. The core problem in open AI ecosystems is determining which agent capabilities (skills) are safe to run. Relying on centralized blacklists is slow and vulnerable. ClawGov shifts this power to the community through a four-step pipeline: First, the system Scans external skill registries like Moltbook to flag potentially risky code. Next, it allows users to Propose governance actions (e.g., banning a specific, high-risk scraping tool). The community then Votes on these proposals using SURGE-backed quadratic voting, which ties governance weight to network participation and ensures fair, Sybil-resistant outcomes. Finally, and most importantly, the system Enforces the community's will. Once a vote passes, ClawGov compiles the decision into a strict policy file deployed directly to the OpenClaw Gateway. If an agent attempts to execute a banned skill, the Gateway hook identifies the unauthorized action and blocks it at the network layer. ClawGov bridges Web3 incentive models with cutting-edge AI infrastructure, proving that decentralized communities can successfully govern autonomous AI behaviors in real-time.

DeMemo

DeMemo is a memory layer in AI that is an enterprise-scale, decentralized, artificial intelligence that is aimed at redefining the manner in which AI systems store, own, and reuse knowledge. It is based on the Arc Network blockchain as an infrastructure, introducing identity management through wallets, where users and applications have complete ownership and control over their data. The system implements end-to-end encryption on all conversational and contextual memory, which guarantees privacy and security on all levels. DeMemo uses permanent and decentralized storage with the help of IPFS but indexes all memory records immutably on-chain. This architecture ensures the persistent storage of data in the long term, verifiable integrity, censorship resistance, and easy portability of AI memory across applications, platforms, and ecosystems without depending on the centralized intermediaries.

Crew-7

Crew-7 is a next-generation multi-agent AI platform built to replicate the structure, behavior, and efficiency of real-world teams across every domain not just software development. Each Crew consists of seven AI agents: 1 Orchestrator and 6 specialists who collaborate using a hybrid communication model inspired by modern engineering, research, business, and creative teams. The Orchestrator plans missions, delegates tasks, and resolves ambiguities, while specialists communicate directly to solve dependencies. This creates a high-performance autonomous workforce capable of tackling complex, multi-step projects with speed and accuracy. Users can choose from prebuilt Crews such as Backend Builder, Frontend Builder, SaaS Architect, Product, Marketing, Research, or Business Analyst or create their own custom team. When a mission is launched, Crew-7 executes inside a secure sandbox, generating complete deliverables: system architectures, APIs, websites, documents, analyses, strategies, and production-grade assets depending on the domain. A real-time Agent Graph visualizes how agents think, plan, and communicate, giving users full transparency into orchestration, reasoning steps, and tool usage. Every event agent start, agent message, tool call, and artifact creation is captured for traceability. Crew-7 provide a built-in Marketplace allows users to install ready-made crews instantly, while the Web3 upgrade layer introduces digital ownership: crews can be minted, verified, traded, rented, or monetized as evolving AI assets backed by on-chain reputation. By combining orchestrated intelligence, domain specialization, real-time visualization, secure execution, and digital ownership, Crew-7 transforms AI from a single assistant into a coordinated, multi-skilled workforce capable of delivering full projects from concept to production across any industry.

Paper2Starup

Thousands of research papers are published every day, but most never reach the real world because researchers lack the tools to turn theory into business. Our project, AI Research-to-Startup Generator, bridges this gap. Using GPT-5, the app takes any research paper (PDF or text) and converts it into a clear plain-language summary, identifies practical startup use cases, generates a draft pitch deck outline, and proposes monetization strategies. Designed with a simple Gradio-based UI, it enables students, researchers, and innovators to instantly explore the entrepreneurial potential of their ideas. Scalable across industries, this tool has strong appeal for universities, innovation hubs, and investors.