Hi everyone,
I’ve been working on define-kit, an open-source TypeScript toolkit for deterministic DeFi position evaluation.
The first position module is for Aave V3 Health Factor.
The main idea is to separate protocol reads from risk evaluation:
fetch → snapshot → evaluate
The adapter reads the position directly from Aave V3, creates a snapshot tied to blockNumber and blockHash, and then evaluates that snapshot without additional RPC calls.
This makes the evaluation deterministic and allows the same snapshot to be stored, compared over time, or evaluated again later.
The current result classifies the position as:
healthy / warning / danger / liquidatable
GitHub:
npm:
https://www.npmjs.com/package/@define-kit/position-modules
I’d appreciate feedback from people familiar with Aave integrations:
- Does the fetch → snapshot → evaluate approach make sense for monitoring Aave positions?
- Are there Aave-specific edge cases or position states that you think the module should handle differently?
The module is MIT licensed and intended to stay reusable and independent of any particular application.