What problem does it solve?
Writing insights discovered during AI-assisted research conversations are easily lost, causing terminology drift and inconsistent style across grant proposals and papers. This Skill captures those insights and merges them into an existing project guide file without restructuring it.
Core Features & Use Cases
- Atomic insight extraction: Splits user input into single-claim statements and maps each to the correct guide section (writing philosophy, terminology, methodology, risks, appendix).
- Minimal-patch updates: Adds only 1-3 bullets into existing sections without creating new level-2 headings or reordering content.
- Structure validation: Runs scripts/validate_guide.py to confirm the seven required headings remain intact after every update.
- Use Case: After a literature review session, ask the AI to distill the findings into your 项目指南.md so terminology and review-response wording stay consistent across the whole proposal.
Quick Start
Update the project guide at projects/MyProject/项目指南.md with the new rule that all abbreviations must use the full-name-plus-abbreviation format on first occurrence.