ChineseResearchLaTeX is a collection of LaTeX templates and an AI-assisted workflow for preparing Chinese research documents such as grant proposals, papers, theses, and academic CVs. Researchers use it to plan, format, review, compile, and revise these documents with human oversight. The catalogue skills and instructions support its agent-based research-writing workflow.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add huangwb8/ChineseResearchLaTeX --skill research-ideagit clone --depth 1 https://github.com/huangwb8/ChineseResearchLaTeXWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/huangwb8/chineseresearchlatex/research-idea)<a href="https://agentmods.dev/skills/huangwb8/chineseresearchlatex/research-idea"><img src="https://agentmods.dev/badge/skills/huangwb8/chineseresearchlatex/research-idea/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/huangwb8/chineseresearchlatex/research-idea"><img src="https://agentmods.dev/badge/skills/huangwb8/chineseresearchlatex/research-idea.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00115 | $0.05178 |
| Opus 5 | $0.00057 | $0.02589 |
| Sonnet 5 | $0.00023 | $0.01036 |
| Haiku 4.5 | $0.00012 | $0.00518 |
Grade A, and why
research-idea scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research Idea
目标
把资料转化为可追溯的研究脉络 map 和科学问题,分开判断科学价值、判断可信度与近期投入,形成可推荐、当前范围内无合格候选或证据不足的结论。候选生成不得脱离前置文献调查,不能由 AI 仅凭资料臆造。
与相邻 skill 的边界:
research-topic-extractor:只负责把资料提炼成可检索主题。research-literature-radar:先发现并筛选经典、前沿和重要论文,形成候选文献池。research-literature-interpretation:逐篇解读入选论文,提取问题、机制、证据、边界和可迁移启发。research-literature-review:负责 Premium 查新和证据综述。parallel-vibe:负责默认 3 轮串行独立审查与打磨。research-plan:在已有科学问题和假设后,才用于实验设计或分析计划。
流程
输入
- 必需:任意资料或信息,如文本、文件、文件夹、URL、论文线索、实验现象、代码仓库或 PR 背景。
- 可选:
- 输出路径:用户指定时遵从;未指定时放在
./docs/ideas/。 - 工作区:用户指定时遵从;未指定时为当前工作目录下
.bensz-api/task-{yyyymmdd-hhmm}-{简短描述}/research-idea/。 - 轮次:默认 3 轮;用户指定时遵从。
执行步骤
初始化与资料归纳
- 先读 运行说明,核对
config.yaml.dependencies.kernel的环境要求。直接用bsk workspace init --task-root复用本轮已声明任务目录,再用scripts/init_workspace.py初始化研究参数和候选空模板、bsk state transition --skill-root进入 literature。已有任务按运行说明读取 Kernel 快照与来源恢复。 - 读取资料,在现有输入摘要中说明目标研究贡献、服务的问题或决策,以及时间与资源约束。资源区分已具备、明确没有、尚不清楚;源码没有某能力不等于团队无法建设。目标未说明时给出暂定解释,只有不同解释会改变主线选择时才澄清。只保存脱敏摘要和必要引用。
- 用
research-topic-extractor生成主题、5-10 个英文关键词、2-5 个核心问题;保存为本 Skill 的input/theme.json(引用主题提取 Skill 的原始产物),字段为topic、keywords、core_questions。
文献调查与解读(候选生成前置)
- 调用
research-literature-radar,根据input/theme.json获取领域内重要、经典、前沿和具有启发性的论文。优先获取公开 PDF 正文;若无法获得 PDF,允许使用题目、摘要和可核验元数据,但必须标记证据深度不足。 - 将雷达结果及其 provenance 保存到本任务
research-literature-radar/output/,至少记录论文稳定 ID、题目、年份、来源、PDF/摘要可用性、入选理由和未覆盖风险。雷达失败或没有达到最低证据量时,不得直接生成候选,应先报告并停止后续依赖步骤。 - 对入选论文调用
research-literature-interpretation,采用并行子 agent 分批执行:- 一个子 agent 只负责一篇论文,独立读取该论文的 provenance 与可用正文/摘要,并将结果写入本任务
research-literature-interpretation/output/下独立的论文目录。 - 同时运行的解读子 agent 最多 3 个(不含负责调度与汇总的主 agent);入选论文超过 3 篇时按批次排队,上一批全部完成(或记录失败)后再启动下一批。
- 本阶段不再嵌套启动额外的并行解读 agent;若单篇需要补证据或定向复核,由该子 agent 在自身任务内完成,不能突破全局并发上限。
- 主 agent 汇总所有成功解读,并保留每篇论文的失败/证据不足状态;任何论文未完成时不得把研究脉络 map 标记为完整。
- PDF 可用时优先基于全文;只有摘要时,解读必须收缩到摘要支持的范围,不得补写全文结论。
- 一个子 agent 只负责一篇论文,独立读取该论文的 provenance 与可用正文/摘要,并将结果写入本任务
- 主 Agent 基于全部解读建立本 Skill 的
output/research-map.md,先读 研究综合指南。主体包含研究线比较表、关系与演化说明、带稳定 O 编号的研究机会清单,并保留时间线与论文 R 编号锚点。区分有来源的关系与待验证综合判断;不能把时间先后、相似术语或不同测量结果写成继承、因果或矛盾。没有争议也可从重要未测量现象与适用边界提出问题,不编造冲突。缺少关键证据时定向补读。
What ships with it
21 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- CHANGELOG.md 7.6 KB
- config.yaml 4.6 KB
- README.md 9.0 KB
- references/agent-review-prompt.md 2.2 KB
- references/novelty-check.md 3.2 KB
- references/report-template.md 6.0 KB
- references/research-synthesis.md 4.4 KB
- references/runtime-guide.md 16 KB
- references/states/candidates/STATE.md 3.0 KB
- references/states/completed/STATE.md 1.8 KB
- references/states/index.json 1.9 KB
- references/states/literature/STATE.md 2.4 KB
- references/states/reporting/STATE.md 3.1 KB
- references/states/review/STATE.md 3.0 KB
- references/verifiers/hypothesis-merit/VERIFIER.md 5.5 KB
- references/verifiers/index.json 1.2 KB
- references/verifiers/stage-readiness/VERIFIER.md 7.2 KB
- scripts/check_completion.py 13 KB runs code
- scripts/check_dependencies.py 3.1 KB runs code
- scripts/init_workspace.py 8.2 KB runs code
- scripts/validate_report.py 14 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today Changed · -8 lines · -28 tokens per session ff3825005df7
- 3d ago Changed · +29 lines · +12 tokens per session dd0f936ed1c6
- 7d ago Changed · +2 lines · -49 tokens per session 6a4b4afb4ed1
- 12d ago First seen · 170 lines · 180 tokens per session scan A de8eb3883bfd
research-idea is a skill published in the GitHub repository huangwb8/ChineseResearchLaTeX (2,726 stars, last pushed yesterday), licensed MIT. It adds 115 tokens to every session and 5,178 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
results-analysis
This skill should be used when the user asks to "analyze experimental results", "run strict statistical analysis", "compare model performance", "generate scientific figures", "check significance", "do ablation analysis", or mentions interpreting experiment data with rigorous statistics and visualization. It focuses on…
nature-data
Prepare, audit, or revise Nature-ready Data Availability statements, data repository plans, dataset citations, and FAIR metadata checklists for manuscripts. Use when the user asks about Nature data availability, research data sharing, repository selection, accession numbers, restricted or sensitive data, source data…
nature-response
Draft, audit, or revise point-by-point reviewer response letters for Nature-family manuscript revisions. Use when the user provides reviewer comments, editor decision letters, revision notes, response drafts, or asks how to respond to major/minor revision requests, rebuttal letters, response to reviewers, peer-review…
paper-self-review
This skill should be used when the user asks to "review paper quality", "check paper completeness", "validate paper structure", "self-review before submission", "audit claims", "check overclaiming", "verify whether results support claims", or mentions systematic paper quality checking. Provides comprehensive quality…
post-acceptance
This skill should be used when the user asks to "prepare conference presentation", "create presentation slides", "design poster", "make academic poster", "write promotion content", "create Twitter thread", or mentions post-acceptance conference preparation. Provides comprehensive workflow for presentation, poster, and…
results-report
A workflow for turning completed experiment analyses into a structured research report with findings, limitations, failures, and next steps.