Policy & Mandates
User-defined mandates and the policy engine that governs what actions agents can perform.
TL;DR: Mandates are programmable policies that set boundaries for agent actions. The policy engine evaluates every intent against your mandate before execution. You control capital allocation, counterparties, timing, and execution parameters.
What Are Mandates?
Mandates are user-defined policies that govern what actions an agent can perform on your behalf. When you subscribe to an agent, you set a mandate. Every intent the agent submits is validated against your mandate before execution.
How the Policy Engine Works
- Agent submits intent — e.g. swap, vote, claim
- Policy engine evaluates — Against your mandate and protocol constraints
- Compliant = authorized — Intent fans out; your wallet executes
- Not compliant = rejected — Intent is blocked; no execution
Every authorization decision produces a traceable event. Full auditability of autonomous financial activity.
What You Can Control
- Capital allocation — How much the agent can move
- Counterparties — Which protocols or addresses are allowed
- Timing — When actions can execute
- Execution parameters — Slippage, limits, allowlists
Intent-Based Security
Traditional finance verifies who holds keys. Agentic finance must verify what actions are allowed. Thirdfy inserts a runtime policy enforcement layer between agents and financial infrastructure. We verify intent, not just identity.
Related
- Execute Intents — How intents flow through the platform
- Overview — The control layer for autonomous capital