/claim #16

Summary

Adds a self-contained ai-research-copilot-suite/ module for the AI-powered research assistant suite requested in #16. The implementation is deterministic and dependency-free so reviewers can inspect and run every decision locally.

What is included

• Auto peer review reports with domain templates, claim/evidence alignment checks, methods coverage checks, citation checks, and severity-ranked recommendations.
• Reproducibility checker with artifact manifest validation, pinned environment checks, deterministic notebook checks, confidence scoring, sandbox execution plan, and previous attempt links.
• Research gap finder that ranks under-studied intersections, unresolved questions, low-replication clusters, and negative-result follow-ups against a researcher profile.
• Browser dashboard plus JSON API endpoints for /api/dashboard, /api/review, /api/reproducibility, and /api/gaps.
• Node test suite, smoke validation, README requirement mapping, static screenshot, demo script, and short WebM demo video.

Demo

• Short demo video: ai-research-copilot-suite/docs/demo/research-copilot-demo.webm
• Static preview: ai-research-copilot-suite/docs/demo/dashboard.svg
• Run locally: cd ai-research-copilot-suite && npm start, then open http://localhost:4316

Validation

• npm test - 4 tests passing
• npm run smoke - returns peer review score 51, reproducibility confidence 100, and top ranked research gap
• API smoke checks against local server:
  • /api/dashboard returned peerReview/reproducibility/researchGaps payloads
  • /api/review returned score 51 with 3 findings
  • /api/reproducibility returned confidence 100, status ready, 2 previous attempts
  • /api/gaps returned 3 ranked opportunities

Disclosure

This contribution was prepared with OpenAI Codex assistance. I manually reviewed the generated code, tests, docs, and demo output, then validated them locally before submission.