National Grant Vetter

Problem

National or provincial education grant materials often show mismatched promises across stages, unsupported data, formulaic wording, or decorative citations. The skill turns that self-review into a concrete checklist for Chinese-language education research projects, covering proposal, approval, mid-term report, and final report.

How it works

The skill first validates input. It accepts a local folder, URL, or pasted text, then infers material type and project level. It then checks six dimensions: lifecycle coherence, authenticity, feasibility, AI-generated traces, fact-checking, and compliance. When multiple documents are provided, it compares them across stages, such as whether the approved scope shrank, whether mid-term outputs match the plan, and whether final deliverables fulfill commitments. AI-trace review looks at sentence uniformity, connector density, missing concrete details, inflated abstract wording, and overly neat paragraph structure; citation review spot-checks key references, policy documents, honors, and internal contradictions. Results are ranked with 🔴, 🟡, and 🟢 risk levels, with actionable revision suggestions.

Boundaries

This is an auxiliary self-review tool for Chinese education-related grants, not an official review. Precision drops for English projects, non-education grants, image-only scans, or outline-only material. AI detection only flags stylistic patterns and cannot establish academic misconduct. Citation and fact checks rely on publicly searchable information, so unpublished data, internal records, and confidential materials cannot be verified.