Commercialize Academic Research

Quick start

1. Require a bounded research_asset before doing substantive commercialization work. If missing, ask for the paper, abstract, prototype summary, deck, technical memo, or lab notes.
2. Default to fast triage unless the user asks for a full case pack, deep analysis, red-team review, validation sprint, or source-grounded market scan.
3. Use this required chain before recommending anything: research claim -> workflow pain -> current workaround -> buyer/budget -> why now -> evidence -> weak link -> validation test.
4. Choose the evidence posture early: user-facts-only, light desk research, or source-grounded diligence. If current industry, regulatory, reimbursement, funding, pricing, or competitive facts materially affect the answer and live research is available, verify them and cite sources; if not available, label them as assumptions.
5. Initialize a working pack with python3 scripts/init_case_pack.py <case-name> --path <output-dir> when the user wants persistent artifacts.
6. For persistent packs, create <case>.source-log.md if missing with python3 scripts/add_source_log.py <case-dir>, maintain it, and validate the pack with python3 scripts/validate_case_pack.py <case-dir> before returning it.
7. Compare at least two plausible commercialization paths before recommending one; do not default to startup formation.
8. End every pass with assumptions, evidence strength, weak links, cheapest next tests, and a kill / pivot / continue decision.

Modes

Fast triage

Use by default for quick commercialization judgment. Produce a 1-2 page chat answer with the required analysis chain, evidence table, path comparison, and top tests. Use live sources only for facts that would materially change the recommendation.