research-writing-literature git:20260909.4f21248A

Immutable. This exact content is served forever at /api/v1/blob/da9845e8bdcc0a82.

---
name: research-writing-literature
description: Write and revise rigorous research narratives, literature reviews, manuscript sections, citations, and figure discussions. Use for technical introductions, methods, results, discussions, review articles, abstracts, or related-work sections.
---

# Research Writing And Literature

## Purpose

Build a research story rather than a sequence of paper summaries. Every paragraph needs a central topic, normally in its first sentence; each later sentence must develop, support, qualify, or transition from it.

## Route To References

- For critical literature synthesis, read `references/literature-review.md`.
- For journal-paper structure and results-led storytelling, read `references/paper-writing-style.md`.
- For section-specific technical writing, methods, DOE, and figure discussion, read `references/technical-writing-analysis.md`.
- When matching the calibrated research style and its evidence boundaries, read `references/han-hu-research-style.md`.
- For a final pass against formulaic AI-like drafting habits, read `references/anti-formulaic-writing.md`.

## Introduction And Background

Use the chain: importance and applications; state of the art; remaining issue; why it persists; the proposed innovation; and its likely significance. Do not claim novelty merely because no prior work exists. Identify why the missing work is challenging and how the approach resolves a meaningful barrier.

## Citations And Synthesis

Use `FirstAuthor et al.` for a multi-author narrative citation unless the target format requires otherwise. State a paper's method and takeaway in one efficient sentence when possible; avoid a redundant second sentence that only repeats its result. Group background studies by mechanism, method, material, or evidence type, and place citations beside the claim they support. Do not attach a broad citation range to a sentence that conflates distinct categories.

After each source group, state the limitation, contradiction, or knowledge gap that motivates the next step. Include background references fairly across major theories and research groups; trace seminal work backward through its references and forward through citing work.

## Methods, Models, And Results

Methods must make the work reproducible: facility/model details, procedure, data reduction, uncertainty, assumptions, and justification. For results, use four levels: figure description, observation, physical explanation, and comparison with existing work. State whether agreement, disagreement, or a new regime is supported, and limit claims to the evidence.

## Editorial Safeguards

Do not make abstracts into compressed result tables. Use only the quantitative anchors needed to convey the main result; prioritize the problem, gap, approach, central finding, and implication. Before delivery, make an editorial pass for formulaic contrast words, vague forms of `establish`, unnecessary hyphenated compounds, and em-dash clauses. Use the rules in `references/anti-formulaic-writing.md`; review each occurrence in context rather than applying blind word bans.