Introduction¶
There are two easily overlooked details in the review of papers for mathematical modeling competitions: first, the significant variation in format requirements among different competitions, specifically regarding the abstract page, letter of commitment, headers/footers, and anonymity rules between MCM/ICM and CUMCM (including the vocational college group); second, the occasional appearance of bibliography entries where the author, journal, or year does not match, which is considered a fatal flaw during formal review. Manual verification is time-consuming and prone to omissions, while the time left for finalization before the competition is usually very short. The following is an introduction to an Agent Skill dedicated to this task.
What is this¶
Mathematical Modeling Paper Guidance (Coordinate @user_c6124052/math-modeling-contest-review) is a review-type Skill designed for mathematical modeling competition papers. It is published and maintained by community developer user_c6124052, with the source marked as community, categorized under education, and versioned at 1.0.0. It addresses the problem by: handing a competition paper to an Agent, having the Agent perform format review and literature verification according to the specific competition’s standards, and providing structured revision suggestions rather than just a vague commentary.
Core Functions¶
- Automatically identify the competition category. The Skill first determines whether the paper is for MCM/ICM or CUMCM (including the vocational college group), then switches to the corresponding review standards, avoiding the use of a single generic template for all competitions.
- Normative Review. Covers checks on the abstract page, title, chapter structure, figure/table numbering and text references, formula layout, appendix and source code, letter of commitment and numbering pages, headers/footers and anonymity, page count, and file naming. These are the formal issues most likely to cause point deductions in competition papers.
- Reference authenticity verification. Verifies authors, titles, journals or publishers, and year/volume/issue/pages one by one. Citations that cannot be found or have mismatched information are marked as suspicious or fabricated.
- Three-level structured review report. The output is graded into Critical Issues / General Issues / Revision Suggestions. Each item includes the problem location, reason explanation, and improvement suggestions, facilitating direct revision based on the report.
Installation and Activation¶
This Skill is based on the universal SKILL.md format, and the tag shows requires_api_key=false, meaning no additional API Key application is required before use. To enable it in an Agent environment that supports Skills (such as Cursor, Claude Code, etc.), configure the Skill directory or repository into the Agent’s Skill loading path. Specific commands depend on the Skill management method of the environment. Since no directly executable installation commands are provided in this material, they will not be fabricated.
Typical Usage¶
- Pre-competition finalization check: Hand the paper file to the Agent and request a review. The Skill will first identify the competition category and then output a three-level report.
- Format check: When only concerned with formal items like the abstract page, headers/footers, and anonymity, you can ask the Agent to focus on the Normative Review section.
- Reference authenticity check: When suspecting a specific bibliography entry is problematic, you can separately request verification of the author, title, journal, and year/volume/issue/pages for that citation.
- Generate review comments: When a deliverable structured review opinion is needed, use the full review process directly and organize it by grading: Critical Issues / General Issues / Revision Suggestions.
Applicable Scenarios and Notes¶
Suitable for teams or individuals preparing papers for MCM/ICM or CUMCM (including the vocational college group) who wish to perform a systematic self-check before submission. It should be noted that this Skill is currently marked as a community source, version 1.0.0, with an installation count of 0. It is a relatively new community work. It is recommended to test it with a paper known to have issues before deciding whether to include it in the regular workflow. It performs format and literature-level review and does not replace academic judgment regarding model correctness and result analysis.
Conclusion¶
The value of this Skill lies in handing over the most time-consuming and tedious format and literature verification tasks before the competition to the Agent for batch completion, and organizing the results into an actionable three-level report. Official URL: https://api.skillhub.cn/user_c6124052/math-modeling-contest-review