[codex] Add domain review template selector

[swhan0329]

Summary

• Add a dependency-free domain review template selector for the AI-Powered Research Assistant Suite.
• Classify manuscripts into clinical, machine-learning, wet-lab biology, materials/chemistry, computational, or general-science review templates.
• Attach claim/evidence alignment, required artifact checks, reproducibility readiness, reviewer expertise routing, blockers, warnings, and a deterministic audit digest.
• Include README, requirements map, acceptance notes, tests, CLI demo, SVG preview, and a short H.264 demo video at domain-review-template-selector/demo.mp4.

Validation

• node domain-review-template-selector/test.js
• node domain-review-template-selector/demo.js
• node --check domain-review-template-selector/index.js
• node --check domain-review-template-selector/test.js
• node --check domain-review-template-selector/demo.js
• git diff --check
• ffprobe confirmed domain-review-template-selector/demo.mp4 is H.264, 1280x720, yuv420p, 5.0 seconds, 150 frames.

Bounty

/claim #16

AI-assisted with OpenAI Codex; reviewed and verified locally before submission.

[swhan0329]

Fresh local verification for review/reward consideration on /claim #16:

• node domain-review-template-selector/test.js -> passed
• node domain-review-template-selector/demo.js -> passed; demo reports status ready, domain machine-learning, template machine-learning-peer-review, digest 0adf94e0597127fe...
• node --check domain-review-template-selector/index.js -> passed
• node --check domain-review-template-selector/test.js -> passed
• node --check domain-review-template-selector/demo.js -> passed
• git diff --check -> passed
• ffprobe domain-review-template-selector/demo.mp4 confirms H.264, 1280x720, yuv420p, 5.0s, 150 frames

The change remains scoped to the domain review template selector slice described in the PR body.