# Glossary

* **Adversarial AI Agents:** AI models trained to simulate real-world attack scenarios, probing smart contracts, tokens, and AI agents for vulnerabilities.
* **Swarm Intelligence:** The collective behavior of autonomous agents working together to simulate complex and evolving threat landscapes.
* **Reinforcement Learning (RL):** A machine learning paradigm where agents learn optimal attack and defense strategies by trial and error, continuously improving based on feedback.
* **Onchain Economy:** The ecosystem of blockchain-based assets, smart contracts, tokens, and decentralized applications.
* **Agentic Web:** The emerging network of autonomous AI agents interacting within decentralized environments.
* **Prompt Injection:** A type of adversarial input manipulation that causes AI agents to perform unintended or harmful actions.
* **Rug Pull:** A fraudulent event where project creators drain liquidity or funds, leaving token holders with worthless assets.
* **Exit Scam:** Similar to a rug pull, involving a sudden shutdown or disappearance by project operators, leaving users without recourse.