Add four focused agent evaluation paper notes

[estelledc]

Summary

• summarize the previous 40-paper round gaps: static-only receipts, short cards, toy-only L4, and missing coverage for MLE / terminal / long-context / visual web agents
• add four study-v2 paper notes: MLE-bench, Terminal-Bench, RULER, and VisualWebArena
• add matching study-review-receipt-v1 files and refresh atlas/count copy

Verification

• node scripts/quality-gate.mjs for each new note: pass, no advisories
• npm run audit:content-contract
• npm run atlas
• npm run audit:counts
• npm run audit:links
• npm run audit:wikilinks
• git ls-files -co --exclude-standard -z | node scripts/audit-public-redlines.mjs --stdin0
• npm run build:strict -- --log /tmp/study-one-more-round-build-clean.log
• STUDY_CHANGED_FROM=origin/main npm run verify:ci

Evidence boundary

All four notes remain STATIC_ANALYSIS / UNVERIFIED. No benchmark harness, Kaggle task, terminal task, RULER run, or VisualWebArena trajectory was executed.