Turn a rough research idea into a structured, evaluable research plan.
Draft, compress, or reformat a manuscript for a specific conference's page limits.
Fact-check claims and citations before submission.
Write a rebuttal and track revisions across review rounds.
| mikubaka88/ccfa-skills | madnanrizqu/vibe-cv-resume | siriusfzh/novaforge | |
|---|---|---|---|
| Stars | 133 | 72 | 199 |
| Language | TeX | TeX | TeX |
| Setup difficulty | moderate | — | moderate |
| Complexity | 3/5 | 2/5 | 2/5 |
| Audience | researcher | general | general |
Figures from each repo's GitHub metadata at analysis time.
Requires initializing the shared ccfa.yaml project state file and reading the venue guide before writing.
This repository contains a collection of structured AI agent skills designed to help researchers write and manage academic papers targeting CCF-A venues. CCF-A refers to the top tier of computer science conferences and journals as ranked by the China Computer Federation, covering events like NeurIPS, CVPR, and similar prestigious publication venues. The core idea is that writing a high-quality paper is not a one-shot text generation task but a continuous process with distinct stages that should not be mixed together. The project organizes that process into 13 separate skills, each with a clear responsibility: one for shaping an early research idea, a separate one for critically evaluating and scoring ideas, another for literature search, another for experiment design, another for writing the manuscript, another for peer-review-style critique, another for fact-checking claims and citations, another for checking submission requirements, and so on through revision tracking and resubmission planning. A shared project state file called ccfa.yaml records the current stage, target venue, which claims have been made, which experiments support them, and what reviewer feedback has been received. Skills read from this shared file but each writes only to its own designated outputs, so the manuscript, the review report, the integrity audit, and the rebuttal each have a clear owner and do not overwrite each other. The README explicitly describes which skill to use for which situation and which skill not to use, because many of these tasks look similar on the surface. For example, refining a vague idea and scoring two competing ideas are handled by different skills. Searching for new supporting literature and verifying whether already-cited papers actually support a claim are also handled separately. The README is written in Simplified Chinese with English and Traditional Chinese versions linked. The project is built using TeX and is intended for researchers working toward publication in competitive academic venues.
A set of 13 coordinated AI agent skills that guide researchers through every stage of writing a CCF-A conference paper.
Mainly TeX. The stack also includes TeX, YAML.
The README does not state a license.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly researcher.
This repo across BitVibe Labs
Verify against the repo before relying on details.