University Faculty Grant Topic Expert

Problem

Faculty grant proposals often stall at the topic stage: candidates cannot find a defensible title aligned with national or provincial priorities, cannot tell whether similar projects have already been funded, struggle to parse detailed call requirements, and produce titles with weak structure.

How It Works

The skill turns topic selection into an auditable workflow:
- Policy scan: extract keyword patterns from national strategy, provincial plans, and call documents;
- Award analysis: inspect the last 3 years of funded projects to map hotspots, gaps, and competition;
- Literature review: use materials from CNKI, Google Scholar, and similar sources to identify research gaps;
- Topic generation: produce 15 candidate titles, score them on policy fit, novelty, feasibility, and related dimensions, shortlist the top 3, and expand them into 24 sub-topics across 6 categories;
- Reporting: generate an HTML justification report and a mock review report.

It can run helper scripts such as policy_scanner.py, project_analyzer.py, and title_generator.py, or work through the four-step dialogue: policy, awards, literature, and topic generation. It is most useful when a call notice, researcher profile, and award list are available.

Boundaries

Without source materials, outputs are generic. Academic database retrieval may require institutional accounts or manually exported papers. Final feasibility still depends on the researcher's prior work, team capacity, and regional competition.