Extend GHO Steward on Aave Prime Instance

The proposal follows the standard Aave governance pattern: the on-chain action is a single execute() call to a dedicated payload contract, which internally performs the detailed steps (granting RiskAdmin to the newly deployed GhoAaveSteward) described in the prose. The decoded calldata here does not expose any internal calls or arguments beyond the fallback execute(), so there is no way to directly verify the specific addresses, config values, or the RiskAdmin grant against on-chain data from this decoding alone. However, this opacity is expected and typical for Aave payload-based proposals (the referenced GitHub implementation and test files corroborate the described logic), not a sign of tampering or mismatch. No numeric value mismatches, target mismatches, or omitted/extra actions can be identified because the single decoded action doesn't reveal granular parameters to compare against. Given the inability to cross-verify specifics, confidence is low, but there is no evidence of any discrepancy between the description and the visible calldata structure, so the overall assessment is consistent with low confidence rather than a flagged issue.