Guardians

Protocol's pause control procedure

Problem

When issues arise (performance issue, critical bug, malicious attempts, etc.), it is important that Ref Finance has a plan to mitigate (i) the risks of contagion and (ii) the impact of such an event (including potential loss of funds).

Often, turning off the frontend/user interface is not good enough because it does not prevent bots from operating, for example.

Solution

The Guardians can be defined as specific NEAR addresses that have the privilege/ability to pause the main contract (v2.ref-finance.near). They are the key participants of a fail-safe procedure, being able to respond in the event of a specific type of failure.

Event/situation that can be defined as, but not limited to:

• Ongoing attack and/or exploit

• Critical bug identified in production (not yet exploited)

• Release/Deployment causing a potential security vulnerability

Procedure

Prerequisite

The Guardians MUST know how to handle the NEAR Command-Line Interface (CLI).

Process

Process Owner: The Guardian

1. Identify a situation that justifies pausing the contract (v2.ref-finance.near)

2. Double check the facts/situation

3. Inform the other Guardians

4. Pause the contract (v2.ref-finance.near)

5. Inform the Team and the DAO/owner of the contract (ref-finance.sputnik-dao.near)

6. Inform the Community

7. Identify the steps to reactivate the contract (v2.ref-finance.near)

Only the owner of the contract can reactivate it (v2.ref-finance.near)

Finally, anyone can see the list of the Guardians by calling the following view method via CLI:

near view v2.ref-finance.near metadata

More info: https://github.com/ref-finance/ref-contracts/pull/50