FeaturedEnterpriseExplainer · Regulation
One mandate, any agent, any service
Clients set what their AI agents may do. Why that mandate has to hold whichever AI they use and whichever service it touches, and how to enforce it.
7 min readLearn
Guardrails for AI agents
that handle money
For platforms opening their services to AI agents, and for anyone running one: where agent limits belong, how agents lose money, and how to enforce a mandate before anything is signed.
7 min readMore articles
10 articlesEnterpriseRegulation
Custody policy engines weren't built for agents
Custody policy engines approve transfers. AI agents act across trading, staking and payments. What an agent policy must check, and where to enforce it.
Explainer · Risks
AI Trading Agent Guardrails: How to Let an Agent Trade Without Risking Your Account
What can go wrong when an AI agent trades for you, which limits matter, and where to enforce them so a mistake, bug or prompt injection can't get through.
EnterpriseSecurity · Regulation
Exchanges gave agents API keys. Agents still need limits.
Exchanges shipped MCP servers and CLIs for AI agents in 2026, mostly on top of API keys. Why trade-only keys aren't enough, and what agent access needs.
Security · Risks
Is It Safe to Give an AI Agent Your Exchange API Key?
A trade-only key stops an AI agent withdrawing your funds. It doesn't stop bad trades. What can go wrong, what to lock down, and how to enforce real limits.
EnterprisePayments
Agent payment mandates need enforcement, not just signatures
AP2 and x402 let agents pay. A signed mandate says what an agent may spend; something still has to stop it spending more. Where enforcement should sit.
Security · Risks
How AI Agents Actually Lose Money: Prompt Injection, Leaked Keys and Plain Mistakes
Every public case of an AI agent losing money, and what caused it. Prompt injection gets the headlines; leaked keys and plain mistakes have cost far more.
EnterprisePayments · Regulation
Final settlement needs checks before signing
Agent payments in stablecoins are final. No chargebacks, no disputes. Why payment providers should require limits enforced before signing, with proof attached.
Setup
How to Run an AI Trading Agent on Hyperliquid Safely
Set up an AI agent to trade Hyperliquid perps with limits it can't override: API wallets, safe defaults, approving rules, placing orders and closing positions.
Security · Explainer
Why Trade-Only API Keys Aren't Safe Enough for Autonomous AI Agents
Disabling withdrawals stops theft, not bad trades. What trade-only keys, exchange risk settings and signing-level limits each protect against, compared.