Master Academic Paper Writing with an 11-Step Agent Skill

Where this fits

You're a researcher or ML engineer who just finished experiments, and now you face the dreaded writing phase. You ask your agent to draft the introduction, but it returns a generic paragraph that misses the novelty. This skill steps in: it guides the agent through an 11-step workflow, providing LaTeX templates and battle-tested tactics for each section—Abstract, Introduction, Method, Experiments, Related Work, Conclusion, and Supplementary. It turns a vague request into a structured, high-quality draft.

Why agents benefit

• Section-specific structure: The agent gets a clear workflow for each part, so it doesn't wander or produce filler.
• Counterintuitive tactics: It applies principles like “write the abstract last” or “lead with the gap” to strengthen framing and novelty.
• LaTeX templates: Ready-made templates for each section reduce formatting overhead and ensure consistency.
• Focused tool usage: The skill restricts tools to file writing and thinking, preventing the agent from going off-track.
• Clear boundaries: It knows when not to be used (e.g., pre-submission review), avoiding misapplication.

Practical scenarios

Drafting the Introduction from scratch

You have experimental results but no text. The agent uses the skill to outline the introduction: start with the problem, then the gap, then your contribution. It applies the “inverted funnel” tactic and produces a draft that leads with novelty—not a generic survey.

Rewriting a weak Method section

Your method section is dense and unclear. The agent reworks it using the skill’s templates, breaking down the pipeline into clearly labeled steps and adding justifications for design choices. The result reads like a paper from a top conference.

Generating LaTeX for a related work section

You need to position your work against a set of references. The skill’s template helps the agent structure the section thematically, not chronologically, and highlight your contribution’s edge. It also outputs proper LaTeX with correct citations.

Add it to your agent workflow

Install the skill with:

npx skills add CamusGIT/EvoQuant --skill paper-writing

Then, in your agent prompt, simply ask:

Use paper-writing to draft the Introduction for my paper on adaptive learning rates. Focus on novelty framing.

The agent will load the skill, follow its workflow, and write the section to a .tex file.

Compare before adopting

Before committing, compare this skill against alternatives:

• Quality signals: Check if the tactics align with your target venue’s style (e.g., ML vs. HCI).
• Maintenance freshness: Look at the GitHub repo (212 stars, active updates) to ensure it stays current with LaTeX conventions.
• Alternatives: Consider paper-planning for story design or paper-review for pre-submission checks—this skill is for writing only.
• Workflow fit: Does it integrate with your existing agent tools? It only needs file write/read and thinking, so it’s lightweight.

Why it is worth tracking

This skill stands out because it encodes real writing heuristics, not just templates. The 212 stars and active author (CamusGIT) signal community trust. Track it if you frequently write academic papers or build agents that assist researchers. Evaluate it when your drafts need sharper novelty framing or when you want to reduce manual LaTeX setup. For a broader view of such skills, check out OpenAgentSkill for related recommendations.