No Cap

Knowledge base layer of human-verified factual data for AI Agents and decentralized applications

No Cap

Created At

ETHGlobal New Delhi

Project Description

NOCAP creates a trustworthy knowledge base of human-verified facts that AI agents can directly access through our API endpoints and developer SDK. By combining World ID's proof-of-humanhood with economic incentives, we solve the fundamental problem of information reliability in the age of AI.

The platform enables verified humans to submit factual claims and collectively validate them through community voting. Each fact is categorized and tagged to create the most precise knowledge base for AI training and inference. AI agents and developers can query specific categories, search by tags, or access the entire verified dataset.

Key outcomes:

  • Reliable AI Training Data: Clean, human-verified facts for AI model training and context
  • Direct AI Integration: REST API endpoints and SDK for seamless AI agent access
  • Categorized Knowledge: Facts are systematically tagged and categorized for precise queries
  • Decentralized Truth Layer: Community-driven fact verification without central authority
  • Scalable Verification: Automated resolution with human oversight ensures efficiency
  • Economic Alignment: Staking mechanisms reward accurate information and penalize misinformation
  • Sybil-Resistant: World ID guarantees only real humans contribute to the knowledge base

AI agents can query "What are verified facts about renewable energy?" or "Give me tagged facts about blockchain scalability" and receive verified, human-validated responses. This creates the foundation for trustworthy AI systems that need reliable factual information.

How it's Made

Frontend: Next.js 15 World App Mini App with World's MiniKit integration

Authentication: World ID zero-knowledge proofs of personhood verified server-side

Smart contracts: Solidity on World Chain handling voting periods, stake management, proportional reward distribution, and fact resolution

Storage: Hybrid approach - detailed content (title, description, sources, tags) on Walrus, verification state and stake on-chain

API layer: REST endpoints serving categorized facts with metadata, search functionality, and real-time factual data for AI Agents.

Developer tools: SDK that simplifies interaction with the verified knowledge-base.

Economic model: Optional ETH staking for fact creators and voters, proportional reward distribution based on stake and correctness

Data structure: Facts include categories, sources, title, description, verification status, cryptographic proofs

Mobile optimization: Native World App experience with proof of personhood and gas sponsorship for all unchain interactions.

background image mobile

Join the mailing list

Get the latest news and updates