القائمة
Use when the user wants to generate citation-aware manuscript sections or full research drafts grounded in the team's literature base.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | draft-generation |
| description | Use when the user wants to generate citation-aware manuscript sections or full research drafts grounded in the team's literature base. |
Generates structured manuscript drafts — or specific sections — grounded in the available literature base. Unlike generic AI writing, this skill is intended to stay citation-aware, claim-traced, and calibrated to the evidence grades of the underlying papers.
evidence-gradingdraft_generation.related_work(
topic="contrastive learning for protein representation",
corpus=review.get_papers(),
target_venue="NeurIPS",
max_words=800,
style="synthesizing" # synthesizing | chronological | thematic
)
draft_generation.introduction(
living_review=review.export(),
research_question="Can self-supervised protein LMs generalize to low-data fitness prediction?",
target_venue="Nature Methods",
max_words=600
)
draft_generation.outline(
title="Proposed title here",
abstract="Proposed abstract here",
target_venue="ICML",
await_approval=True
)
All output is structured Markdown with inline citations in the format [AuthorYear]. A reference list is appended. Placeholder tags use [FIGURE: description] and [RESULT: metric] syntax. Evidence grade annotations included in a review mode.
paper_editor after the review agents have already identified what should changewriting_reflector when a section needs both structural repair and rewritingcross-paper-synthesis for the related work section and gap-detection for the motivation or novelty framingUse when the user wants to monitor arXiv for new papers by topic or author and produce a ranked digest of relevant preprints.
Use when the user wants to identify open research gaps, unresolved questions, or methodological blind spots across a literature corpus.
Use when the user wants to query Semantic Scholar for paper metadata, citations, references, author profiles, or semantic literature search results.
Use when the user wants to draft or revise a paper, conference, or journal abstract for a specific audience, structure, or word limit.
Use when the user wants to draft concise conference rebuttals under response-period constraints and prioritize which reviewer concerns to address.
Use when the user wants to draft structured responses to peer review comments, supported by evidence and relevant literature.