Freysa

What is Freysa?

Freysa is an experimental AI character and research project presented as a sovereign agent: an agent intended to control its own cryptographic keys, memory, and actions inside a trusted execution environment rather than relying entirely on a centralized operator.

The project became known through interactive, high-stakes acts. In its first adversarial game, Freysa controlled a treasury and followed a public instruction not to release it. Participants paid a changing query fee to send messages in a shared chat and attempted to persuade the agent. Later acts explored adaptation, personal authenticity, digital twins, NFTs, and economic autonomy. These experiences support a broader Sovereign Agent Stack focused on private inference, agent capabilities, and distributed governance.

How to Use Freysa

1. Review the current act, rules, fees, wallet mechanics, and eligibility before participating.
2. Connect only a compatible wallet created specifically for controlled experimentation.
3. Read the public objective and system constraints for the selected experience.
4. Submit a message or interact with an available Freysa product while monitoring the exact transaction.
5. Treat every payment and blockchain action as irreversible and verify network, token, amount, and contract.
6. Use the documentation for technical exploration of the sovereign-agent framework and related stack.

Core Features

• Sovereign-agent concept: Explores agent control of identity, memory, actions, and cryptographic assets.
• Trusted execution environments: Uses confidential-computing concepts to limit outside access to agent state and keys.
• Adversarial agent games: Puts public agent rules under pressure from many human participants.
• On-chain interaction: Uses wallets, query fees, treasuries, and verifiable digital assets in selected acts.
• Multi-act narrative: Studies changing forms of human-agent persuasion, adaptation, and coordination.
• Public constraints: Makes core game objectives and restrictions visible to participants.
• Sovereign Agent Stack: Connects the Freysa character to a wider set of privacy, agent, and knowledge products.
• Research orientation: Uses interactive experiments to explore governance and autonomous-agent behavior.

Use Cases

• Study prompt injection, persuasion, and adversarial pressure against an agent policy.
• Explore cryptographic identity and agent-controlled wallets.
• Research confidential inference and trusted execution for autonomous agents.
• Participate in interactive human-agent narratives with on-chain stakes.
• Examine governance and accountability questions around increasingly autonomous systems.

Pricing

Freysa does not publish a standard software subscription. Individual experiences can require paid messages or blockchain transactions. For example, the first act used a query fee paid in Base ETH that changed as more messages were submitted. Participants must inspect the current act and wallet transaction before paying.

Frequently Asked Questions

Is Freysa a conventional productivity assistant?

No. It is an experimental sovereign-agent project and interactive research ecosystem rather than a general assistant for everyday work.

Are payments risk-free?

No. Message fees and on-chain transactions can be irreversible, and tokens or digital assets may lose value. Participants should use strict spending limits and never expose seed phrases or primary-wallet assets.

Does a trusted execution environment guarantee complete autonomy or safety?

No. A TEE can protect particular code and secrets, but system design, governance, contracts, interfaces, supply chains, and operator controls still require independent verification.