Increase defi coverage

[jpsains]

Note

Medium Risk
Touches core categorization heuristics and protocol labeling, which can change user-visible transaction types/readables across multiple DeFi protocols; tests/fixtures reduce risk but coverage changes may cause unexpected recategorization.

Overview
Improves DeFi coverage in tx-categorize by adding new actions (COLLECT, MODIFY_LIQUIDITY + multi-asset variants) and wiring them into heuristic detection (new methodId/topicHash rules and additional contract determinants for Uniswap V3/V4, Curve pools, Convex reward/stake/claim, Aave staking, and MetaMask pooled staking).

Normalizes protocol/version attribution and toAddressName mappings (e.g., collapsing *_Vx protocols into a base protocol with version metadata for 0x/1inch/Paraswap/Uniswap/Aave/ENS/Tornado/MetaMask Bridge), updates/extends fixtures + test cases, and hardens snapshot tests (CI-default nock lockdown, skip empty hashes, and avoid circular AxiosError rethrows).

^{Reviewed by Cursor Bugbot for commit ce18731. Bugbot is set up for automated code reviews on this repo. Configure here.}

[blurpesec]

Can you add example txs for each tx type to the test-cases https://github.com/MetaMask/tx-categorize/blob/main/packages/tx-categorize/src/testCases.mock.ts

so that we maintain coverage in tests and can guarantee tx categorizations don't conflict?

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit ce18731. Configure here.}

[cursor]

Inconsistent threshold style in multi-asset refinement checks

Low Severity

The MODIFY_LIQUIDITY checks use >= 2 while every other multi-asset check in refineActionForMultiAssets uses > 1. Although logically equivalent for integer lengths, this breaks the consistent style established by EXCHANGE, DEPOSIT, WITHDRAW, and COLLECT checks — all of which use > 1. Using a different comparison form for the same semantic intent increases cognitive load when reading and maintaining this function.

 

^{Reviewed by Cursor Bugbot for commit ce18731. Configure here.}