[wallet_screening] 2.2a: unify TRM/scam tx-screening index + tests (parent: #115)

[Hendobox]

Parent

#115

Scope (Phase 2.2a only)

Build a unified transaction-screening risk index by merging:

• malicious_scs_2025.json
• relevant ETH entries from normalized_uniswap_trm.json (and existing normalized datasets loaded by the skill)

Then use this merged index inside transaction analysis so malicious/risky counterparties are flagged beyond the current 6-contract list.

In scope

• Initialize merged tx-risk index at skill startup
• Normalize/filter to Ethereum 0x + 40 hex addresses
• Use merged index in tx screening logic
• Add tests covering TRM/scam tx interaction detection
• Keep existing tests passing

Out of scope

• Report schema v2 changes
• ERC-20/internal tx (tokentx, txlistinternal)
• Etherscan pagination/truncation metadata
• Docs/card/instructions alignment (Phase 3a)
• Data pipeline/canonical schema (Phase 1)

Acceptance criteria

• A tx involving a known TRM/scam ETH address is detected as a malicious/risky interaction
• Existing sanctions behavior from 2.1 remains intact
• pytest tests/skills/finance/test_wallet_screening.py passes

[Hendobox]

Implemented in PR #140: https://github.com/ARPAHLS/skillware/pull/140\n\nIncludes:\n- unified tx-risk index at init\n- tx analysis lookup against merged index\n- tests for TRM counterparty detection and source merge\n\nKeeps 2.2a scope (no schema v2 / pagination / ERC-20/internal tx).