Agentic Trading with Safe Guardrails

Shuriken has launched a new infrastructure that enables AI agents to perform autonomous trading across all major asset classes, incorporating safety guardrails that restrict their actions based on user permissions. This marks a significant advancement in the development of autonomous finance, allowing agents to act on-chain with granular control.

The Shuriken platform provides the tools and protocols for AI agents to trade in real-time markets, including cryptocurrencies, perpetuals, real-world assets, pre-IPO equity, and prediction markets. The infrastructure emphasizes security by avoiding seed phrases in runtime and offering detailed permission settings, ensuring users retain control over agent actions.

The platform includes a comprehensive integration guide and skills repository, allowing external agents and developers to connect securely via plugins and APIs. These tools facilitate authentication, permission scoping, and API calls, integrating seamlessly with popular coding assistants like Claude, Codex, and Copilot, among others.

Why It Matters

This development matters because it represents a major step toward fully autonomous financial trading, potentially increasing efficiency and market responsiveness. The safeguards aim to mitigate risks associated with AI-driven trading, addressing concerns about security and user control in autonomous systems. If successful, it could reshape how assets are traded and managed in decentralized finance.

Background

Until now, AI agents could perform many tasks but were limited in executing autonomous trades due to security and control concerns. Shuriken’s new infrastructure builds on recent trends toward automation in finance, leveraging on-chain data and real-time information feeds. The platform’s release follows ongoing industry discussions about the safety and regulation of autonomous financial agents.

“We are at the beginning of autonomous finance. Our platform provides the infrastructure for agents to trade securely across all major asset classes with user-defined safeguards.”

— Shuriken team

“The introduction of granular permission controls could be a game-changer for autonomous trading, balancing innovation with security.”

— Industry analyst

What Remains Unclear

It is still unclear how widely adopted the platform will become or how effective the safeguards are in preventing misuse. Details on regulatory compliance and real-world testing outcomes are not yet available, and the security of the permission system remains to be validated in live environments.

What’s Next

Next steps include broader onboarding of external developers and agents, real-world testing of the safeguards, and monitoring of market responses. Regulatory discussions and potential updates to improve security and compliance are expected in the coming months.

Key Questions

How does Shuriken ensure the security of autonomous trades?

Shuriken employs granular permission controls that restrict agent actions based on user-defined scopes, avoiding seed phrases and enabling detailed oversight.

Can users fully control what their agents do?

Yes, users can set specific permissions for each agent, controlling which assets and actions they are authorized to perform.

Is this platform compliant with regulations?

Regulatory compliance details are still emerging; the platform emphasizes security and user control but further legal review is ongoing.

What types of assets can agents trade?

Agents can trade across cryptocurrencies, perpetuals, real-world assets, pre-IPO equity, and prediction markets.