feat: add Breaking Intelligence v2 calibration

[XucroYuri]

Summary

• Adds Breaking Intelligence v2 fields and D1 migration for raw score, percentile, calibrated score, and adversarial flags.
• Introduces rolling calibration for Worker imports/scheduled runs, with changed-row-only event updates to avoid D1 write amplification.
• Aligns public featured/bootstrap/snapshot reads with v2 quality gates and dynamic Breaking polling intervals.

Scope

• Public API URLs and existing JSON fields remain backward compatible.
• This PR does not perform a live deploy and does not claim a production live receipt.
• Source coverage remains validated by the existing source_coverage_report --minimum-refs 20 gate.

Test Plan

• ./.venv/bin/python -m pytest tests/unit/test_breaking_scoring.py tests/unit/test_cloudflare_native_config.py tests/unit/test_cloudflare_refresh_public_snapshots.py -q --tb=short
• cd frontend/public && npm test -- src/lib/feed-state.test.ts src/lib/api.test.ts
• cd frontend/cloudflare && npx wrangler deploy --env="" --dry-run --outdir /tmp/ns-worker-breaking-v2-dry-run --containers-rollout none
• cd frontend/public && npm run lint && npm test -- src/App.test.tsx src/lib/api.test.ts src/lib/feed-state.test.ts src/lib/seo/site-seo.test.ts
• cd frontend/public && npm run build
• cd frontend/admin && npm run lint && npm run build
• ./.venv/bin/python -m pytest tests/unit -q --tb=short
• ./.venv/bin/python tools/source_coverage_report.py --minimum-refs 20 --pretty
• python tools/check_no_hardcoded_target.py
• python tools/scan_sensitive_data.py
• git diff --check

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.