nsfc-abstract_skill

Overview

This skill generates paired Chinese and English NSFC grant proposal abstracts and title suggestions tailored for submission requirements. It produces a five-sentence Chinese abstract (≤400 characters by default) and a faithful English translation (≤4000 characters), plus one recommended title and five candidate titles with reasons. The final output is written to ./NSFC-ABSTRACTS.md and validated against strict length rules when requested.

How this skill works

The skill asks for missing key inputs if the user provides fragmented information, then constructs a five-element abstract: importance, scientific gap, feasibility evidence + hypothesis, 3–4 concise research content points, and expected significance. It creates the Chinese text first, translates it faithfully into English without adding claims, formats title suggestions with reasons, and writes the structured result to NSFC-ABSTRACTS.md. When strict size constraints are requested, the skill runs an automated three-round compress-and-validate loop using the provided validation script until the text meets limits or user guidance is required.

When to use it

• User explicitly requests writing or polishing NSFC proposal abstracts (Chinese + English)
• User requests translating a Chinese abstract into an English abstract for NSFC use
• User needs title suggestions aligned to NSFC title style plus reasons
• When the abstract must meet NSFC character limits and pass automated validation
• Not for general-purpose text translation or drafting other proposal sections (use corresponding nsfc skills)

Best practices

• Provide the info form fields or answer the 3–6 clarifying questions to avoid follow-up
• Give 1–2 quantitative pre-results if available (improves feasibility evidence)
• Specify desired title orientation: mechanism/method/translation/scene if you have a preference
• Request strict validation only when you can accept iterative compression (up to 3 automatic rounds)
• Avoid marketing phrases; supply verifiable claims if you want them retained

Example use cases

• Create a 5-sentence Chinese abstract + faithful English translation for an NSFC application
• Translate an existing Chinese NSFC abstract into an English version that matches the Chinese meaning
• Generate one recommended title and five candidates with concise reasons for selection
• Run strict character-limit validation and compress the abstract until it fits NSFC limits
• Prepare an abstract when you have partial inputs—skill will ask targeted questions to complete missing pieces