ieee-paper-drafter
Skill1 starsChanged 52 days agoMITView on GitHub
Draft an IEEE-conference/journal-style research paper — full section structure, IEEE numeric citation style, and a ready IEEEtran LaTeX skeleton — with the same rigor bar a reviewer would apply. Use whenever the user wants to write up a research idea, project, or result as an IEEE-format paper, or asks for an abstract/related-work/methodology section in that style. Do not use this for informal write-ups or blog-style summaries — that's plain markdown, not this skill.
In this repository32 other files it gives its agents.
- agent-system-designSkill
- api-contract-designSkill
- architecture-designSkill
- architecture-optimizationSkill
- ci-speedupSkill
- code-handoffSkill
- cross-ai-handoffSkill
- crypto-storage-auditSkill
- database-designSkill
- dataset-sourcingSkill
- distributed-debuggerSkill
- embedded-constrained-codingSkill
- freedom-to-operate-analysisSkill
- frontend-architectureSkill
- github-masterSkill
- idea-finderSkill
- memory-leak-hunterSkill
- ml-model-selectionSkill
- ml-security-auditSkill
- mobile-platform-architectureSkill
- observability-designSkill
- patent-drafterSkill
- patent-novelty-evaluatorSkill
- performance-optimizationSkill
- project-scaffoldingSkill
- researcher-evaluatorSkill
- researcherSkill
- safety-compliance-scopingSkill
- scale-refactorSkill
- secure-code-reviewSkill
- slow-query-optimizerSkill
- test-architectSkill
The file
Files
IEEE Paper Drafter
A paper that reads well to its author and a paper that survives peer review are different documents. The gap is almost always: omitted limitations, a related-work section that lists instead of compares, and improvement claims without a stated baseline. This skill exists to close that gap before submission, not after rejection.
Step 1 — Use the template
Copy assets/ieee_template.tex as the starting file (standard IEEEtran conference class, ready to compile). It has the full section skeleton with inline comments on what each section actually needs — follow the section order; it matches reviewer expectations and deviating from it invites confusion, not creativity points. If the user wants a .docx instead of LaTeX, use the docx skill for the file mechanics but keep the same section structure and rigor rules below.
Step 2 — Section-by-section rigor rules
- Abstract: self-contained, 150-250 words, no citations, includes the single strongest result as an actual number — not "significant improvement," but "X% reduction in Y." Every superlative in the abstract must be backed by a metric stated later in the paper.
- Introduction: state the specific mechanism of failure in existing approaches, not "prior work has limitations." List contributions as explicit bullets — a reviewer scanning for contributions should not have to infer them from prose.
- Related Work: must compare mechanisms against the nearest competing work, not just cite and summarize each one in isolation. Run the
researcherskill to find the actual closest work if it isn't already identified, then theresearcher-evaluatorskill against this section specifically — an incomplete or list-style related-work section is the most common reason a technically sound paper gets rejected on presentation grounds. - Proposed Method: enough detail for a reader with no code access to reproduce it. Define every symbol before using it. State non-obvious design decisions and name the rejected alternative and why — this preempts the review comment "why not X instead."
- Experimental Setup: baselines must include the strongest real competitor found during related-work research, not an easy strawman — a paper that only beats a weak baseline gets flagged immediately by any competent reviewer.
- Results and Discussion: report negative and limiting results, not just the wins. Every claimed improvement needs the specific metric and the specific baseline it beat, stated together, not separated across paragraphs.
- Limitations: an explicit section, not a buried caveat. State plainly what the method does not solve and under what conditions it breaks — reviewers penalize omitted limitations far more heavily than disclosed ones.
- Conclusion: restate the contribution against the original problem, not a generic "we presented a novel approach" — that sentence is a reliable signal to reviewers that the author isn't sure what the actual contribution was.
Step 3 — Citation discipline
IEEE style is numeric, ordered by first appearance in text ([1], [2], ...), with the reference list in that same order — not alphabetical. Each reference entry: author initials, title in quotes, venue, year, page range. Never cite a source for a claim stronger than what that source actually supports — this is the single fastest way to lose reviewer trust once caught.
Step 4 — Before calling it done
Run the researcher-evaluator skill against the Related Work and Results sections specifically, since those are where unsupported or miscalibrated claims do the most damage. A paper with a strong method section and a weak related-work section still gets rejected — reviewers judge the comparison, not just the contribution in isolation.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

