skills.rest
zniihgnexy
zniihgnexyCOMMUNITY
Explore all skills by this author →
GitHub

write

Drafts evidence-grounded research papers and reports with LaTeX venue templates and claim-evidence tracking.

Updated Apr 1, 2026
Search Tags:#citations#LaTeX#peer review#paper drafting#research writing#evidence mapping#academic templates
One-click install
npx skills add https://github.com/zniihgnexy/vid_tokenizer --skill write-zniihgnexy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: write
Source: https://github.com/zniihgnexy/vid_tokenizer/tree/main/.ds/codex-home/skills/deepscientist-write
Command: npx skills add https://github.com/zniihgnexy/vid_tokenizer --skill write-zniihgnexy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? Turning experimental results into a faithful research paper is error-prone: claims drift from evidence, outlines lose sync with experiments, and drafts get polished before the supporting data is stable. This Skill enforces an evidence-first writing workflow so papers, reports, and research summaries only state what durable artifacts actually support. ## Core Features & Use Cases - Outline-first paper contract: Maintains a selected outline, evidence ledger, and paper experiment matrix that gate drafting until required evidence is complete. - Venue template support: Ships LaTeX templates for ICLR, ICML, NeurIPS, COLM, AAAI, ACL, OSDI, NSDI, ASPLOS, and SOSP, copied into a paper/latex workspace. - Review and proofing pipeline: Includes reviewer-first writing guidance, five-pass revision checklists, sentence-level proofing, and figure palette discipline for publication-quality output. - Use Case: After finishing a machine learning experiment with an accepted baseline, use this Skill to select an outline, map claims to result artifacts, draft sections in the ICLR 2026 template, run a harsh self-review, and produce a submission-ready paper bundle. ## Quick Start Ask the agent to draft a research paper from the completed experiment results using the write skill with the ICLR 2026 template.

Frequently Asked Questions about write

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

FAQPage Schema
How do I write a research paper from experiment results with AI?▼

Start by consolidating your baseline and experiment artifacts, then select a structured outline that binds each section to required evidence. Draft sections only after the claim-evidence map and experiment matrix confirm main-text items are complete, then run review and proofing passes.

What LaTeX templates are available for ML conference papers?▼

The skill bundles official-style templates for ICLR 2026, ICML 2026, NeurIPS 2025, COLM 2025, AAAI 2026, and ACL, plus USENIX and ACM SIGPLAN formats for OSDI, NSDI, ASPLOS, and SOSP. General ML papers default to the ICLR 2026 template.

When should I not start drafting a paper yet?▼

Do not draft when the evidence base lacks an accepted baseline, when main claims have no durable artifact support, or when feasible experiments in the matrix remain unaddressed. The correct move is to surface the evidence gap and route back to experiments or analysis first.

How do I switch the AAAI 2026 template between anonymous and camera-ready?▼

Uncomment the line \def\aaaianonymous{true} near the top of the unified template for anonymous submission, or comment it out for camera-ready. The template conditionally loads the submission option and adjusts title, links, and content sections automatically.

Why does AAAI LaTeX compilation fail with an illegal bibstyle error?▼

The aaai2026.sty file already sets \bibliographystyle{aaai2026} internally, so adding another \bibliographystyle command in your document causes the error. Remove your own bibstyle line and keep only \bibliography{aaai2026}.

How many references should a research paper draft include?▼

For a normal paper-like deliverable, aim for roughly 30 to 50 verified references gathered through broad literature discovery and close reading. Fewer references are acceptable only when the scope clearly justifies it.

Related Skills

Explore More Education & Research Skills→
OpenLAIROpenLAIR
1.1k

ds-analysis-campaign

Orchestrates coordinated follow-up experiment campaigns for ablations, robustness checks, and error analysis.

Official
Advanced
jandan138jandan138

paper-writing

Generate a complete paper directory with LaTeX sources and PDF from a narrative report.

Community
Advanced
xlinh2301xlinh2301

scientific-figure

Generates publication-quality scientific figures from data through iterative adversarial critique and grading.

Community
Advanced
OpenLAIROpenLAIR
1.1k

ds-write

Drafts evidence-grounded research papers and reports with LaTeX venue templates and citation integrity checks.

Official
Advanced