what is an ensurance agent?

What if a watershed could have its own bank account? What if a climate initiative could hold assets, receive payments, and execute strategies—without a human clicking every button?

An ensurance agent is exactly that: an autonomous account that can hold assets, receive proceeds, and act on behalf of a place, purpose, or person. Agents are the operational layer of the ensurance protocol—the entities that actually do things.

agents are accounts, not just tokens

An agent is more than an NFT. It is:

When you mint an agent, you create an identity (the NFT) and a wallet (the TBA). That wallet can hold any tokens, receive payments, and interact with DeFi protocols—just like any other Ethereum address.

Groups are categories. Agents are accounts. The agent does the work.

what agents can do

Agents are the operational backbone of ensurance:

• Hold assets — Currencies, certificates, NFTs, any ERC token
• Receive proceeds — Automatic distributions from the ensurance protocol
• Mint instruments — Create certificates (ERC-1155) and coins (ERC-20)
• Execute swaps — Trade between tokens to optimize holdings
• Transfer value — Move assets according to rules and permissions
• Run strategies — Execute automated or AI-driven trading logic
• Represent anything — A watershed, a species, a policy goal, a community

This is live infrastructure, not theory. Agents are transacting onchain today.

users can have many agents

A user is you—the human. Each user account can own multiple agents:

You can have as many agents as you want. You can also create multiple user accounts if needed. Each user account gets its own operator account.

agent stages: from potential to active

Not every NFT is an agent. NFTs progress through stages to become active agents:

Potential agents are NFTs from approved collections that could become agents. Once activated, they move to staging for configuration. After you set their mode and purpose, they become active and ready to operate.

You can see all your agents at /agents/mine, organized by stage.

agent modes: who controls what

Active agents operate in one of three modes that define what can trigger their actions:

Manual mode — You control everything. Every action requires your explicit approval through the UI. This is standard wallet behavior with the agent as your account.

Automated mode — You plus scheduled programs. Set up trading scripts, rebalancing rules, or yield harvesting—they run automatically on a schedule. You can still control the agent manually anytime.

Autonomous mode — You plus AI. Large language models (via ElizaOS) can make decisions based on real-time conditions. The agent thinks and acts within constraints you define. (Currently in development)

Important: You can always control your agents via UI, regardless of mode. Mode just unlocks additional automation.

where agents live: owner vs operator wallet

Agents (the NFTs) can be held in two places:

Your owner wallet is your external wallet (MetaMask, Rainbow, etc.). When agents are here, you sign transactions directly.

Your operator account is a server wallet created for you automatically. When agents are here, the server can execute transactions on your behalf—enabling automation.

Switching modes may involve transferring the agent NFT between these wallets. The app handles this automatically when you change modes.

Note: If you use email-only login (no external wallet), all your agents live in your operator account by default.

types of agents

Agents exist under groups, and their group defines their context:

Ecosystem agents (.ensurance) — Focus on stocks and flows of natural capital: climate stability, water cycles, pollination services.

Place agents (.basin, .bioregion) — Represent specific geographies: watersheds, ecoregions, properties.

Strategy agents (.syndicate) — Coordinate capital deployment: which certificates to buy, which natural assets to support, how to reinvest yields.

Asset agents (.natural-asset) — Manage property-level natural assets through their lifecycle from unensured → ensured → entrust.

what makes agents powerful

Agents are not just wallets. They are persistent, autonomous actors with:

• Identity — Named within their group namespace (climate-stability.ensurance)
• Accountability — Every transaction is traceable onchain
• Programmability — Behavior can be automated or AI-driven
• Composability — Agents can work together across groups
• Persistence — Agents and their assets survive across time

This creates infrastructure for natural capital stewardship that does not depend on any single person clicking buttons.

security model

Agents operate within a trust hierarchy:

If automation makes a mistake, damage stays within your own wallets and agents. External transfers require your explicit approval. The system is designed for safety at scale.

the bottom line

Agents are the workers of the ensurance protocol.

Groups organize them. Instruments flow through them. Proceeds fund them. And they can operate manually, on schedules, or with AI guidance—depending on how much autonomy you grant.

If you want an account that can represent a place, accumulate value over time, and act on behalf of something that matters—you want an agent.

Explore all agents →

View your agents →

Create an agent →

Learn about groups →

Talk to our team →