ts-paper-refine

Right-sizes and polishes journal paper drafts while preserving citations, equations, and structure.

1.1k|19|Updated Jun 18, 2026
One-click install
npx skills add https://github.com/Spark-To-Paper-Skills/spark-to-paper-skills --skill ts-paper-refine-spark-to-paper-skills
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ts-paper-refine
Source: https://github.com/Spark-To-Paper-Skills/spark-to-paper-skills/tree/main/skills/ts-paper-refine
Command: npx skills add https://github.com/Spark-To-Paper-Skills/spark-to-paper-skills --skill ts-paper-refine-spark-to-paper-skills

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Drafted academic paper sections often miss journal length targets, repeat content across sections, and carry AI-writing tells, requiring a coherent whole-draft polish pass before peer review. ## Core Features & Use Cases - Length right-sizing: Expands or trims each section to the word bands defined in the template's template.json, targeting a substantial journal article rather than conference compression. - Coherence and de-AI pass: Enforces consistent terminology across sections, removes redundancy, and rewrites AI-tell phrases and comma-soup prose into natural academic writing. - Preservation guarantees: Keeps every citation, equation, reference, table label, and result placeholder intact, with lint scripts verifying the result. - Use Case: After drafting a Traitement du Signal paper, run this stage to bring the method section from 1200 to 2200 words with real substance, scrub tell phrases like "delve into", and verify with draft_lint and citations_lint before handing off to the review stage. ## Quick Start Polish and right-size my drafted TS paper sections to the template word bands, then re-run the draft and citation linters until both pass.

Frequently Asked Questions about ts-paper-refine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I polish an AI-drafted academic paper before submission?

Run a single holistic refinement pass over the full draft that right-sizes each section to the template's word bands, enforces cross-section terminology consistency, removes redundancy, and rewrites AI-tell phrases. Verify the result with automated lint scripts before review.

What word count should each section of a journal paper have?

For the ts_iieta template, targets are roughly 2000-3000 words for method, 1000-1500 for experiments, 800-1200 for introduction and related work, 900-1400 for analysis, 200-280 for conclusion, and 150-220 for the abstract. Other templates define their own bands in template.json.

How do I remove AI writing tells from academic prose?

Replace canned phrases like "delve into", "plays a crucial role", and "it is worth noting", fix comma-soup enumerations with colons or em-dashes, vary sentence length, and synthesize related work instead of listing papers. The draft_lint script hard-fails on context-free canned phrases.

Does refining a paper draft risk breaking citations or equations?

No, the refinement pass preserves every \cite{}, equation, \ref{}, table label, and result placeholder exactly. After editing, citations_lint and draft_lint re-run until both exit with ok:true, confirming nothing broke.

When should I not compress a paper draft to fit page limits?

Do not compress when targeting a journal venue like Traitement du Signal, where articles run 8-14 two-column pages. Compression is only appropriate for conference templates; let the template's word bands, not venue habit, decide length.