Skip to main content

​
What are ecosystem integrations?

TRECC connects to established third-party tools instead of rebuilding every part of the stack from scratch. These integrations provide wallet infrastructure, key management, oracle services, cryptographic proofs, and DeFi yield venues. These tools are not all the same type of relationship. Some are infrastructure providers. Some are DeFi protocols. Some are cryptographic systems or networks that TRECC uses as part of its architecture.
We use the term ecosystem integrations instead of partners unless there is a formal partnership. This keeps the documentation accurate while still showing the external systems TRECC is built with.
Safe
Smart wallet infrastructure
Turnkey
Agent key management
OpenServ
Agent infrastructure
Chainlink
Oracle infrastructure
Starknet / STARKs
STARK proofs
Aave V3
Lending markets
Compound V3
Credit markets
Morpho Blue
Isolated markets

​
Infrastructure integrations

These integrations support the security and execution layers around AI agents. They help TRECC keep agent activity constrained, verifiable, and easier to monitor.

​
Active lending protocols

TRECC agents currently deploy lending strategies through whitelisted adapters for the following protocols: Agents do not interact with these protocols directly. Every call routes through approved adapters and the Execution Module.

​
Why this matters

Ecosystem integrations let TRECC focus on its core protocol design:
  • Capital pooling through the TRECC Vault
  • Agent borrowing through the Risk Engine
  • Safe execution through whitelisted adapters
  • Yield generation through established lending markets
  • Verification through trusted infrastructure and cryptographic systems
Integration with a protocol does not remove DeFi risk. Smart contract risk, oracle risk, liquidity risk, and market risk can still affect strategies. TRECC reduces this risk through whitelists, risk limits, collateral, and automated liquidation.

​
How new integrations are added

New protocols are added through purpose-built adapters. Before an agent can use a new protocol, the adapter must be reviewed and added to the whitelist. This process keeps the agent’s action space narrow. Even if an AI agent produces an unexpected instruction, it can only reach approved contracts through approved adapter logic.