SKILL.md - paper_scout
Overview
INLINECODE0 automates searching and summarizing research papers in robotics and related fields. It uses CrossRef as the primary data source, with Google Scholar as the supplementary source for missing fields or broader coverage.
Functionality
- 1. Primary Query Engine:
-
CrossRef: Quick and structured search for venues like TRO, IJRR, Science Robotics, etc.
- 2. Backup Source:
-
Google Scholar: Scrapes data using real-time browser automation for keywords and specific venues.
- 3. Filters:
- Keywords:
>
muscle parameter estimation,
reinforcement learning,
imitation learning,
human intention prediction,
human-robot interaction collaboration.
- Top venues: TRO, IJRR, RAL, Science Robotics, TMech, CVPR, ICLR, ICCV.
- Date: Year filtering to prioritize recent publications (e.g.,
since 2024).
- 4. Data Outputs:
Structured Markdown reports saved to
~/Desktop/YYYY-MM-DD-academic-digest.md with fields:
- Title
- Source venue (e.g., Science Robotics)
- Year
- Authors
- Abstract
- Major contributions/innovation points
Example Configuration
CODEBLOCK0
Example Usage
- 1. Query Crossref for
reinforcement learning robotics papers since 2024. - Supplement missing information by scraping Google Scholar (
sorted by date). - Remove duplicates, filter results to top academic venues.
- Save a Markdown digest to the desktop.
Notes
- - Automates cross-source integration with duplicate removal.
- Extensible for more specialized sources (e.g., IEEE, PubMed).
- Browser scraping (Google Scholar) requires interactive sessions for JS-heavy pages.
SKILL.md - paper_scout
概述
paper_scout 能够自动搜索并总结机器人学及相关领域的研究论文。它以 CrossRef 为主要数据源，以 Google Scholar 为补充源，用于获取缺失字段或扩大覆盖范围。
功能
- 1. 主要查询引擎：
-
CrossRef：针对 TRO、IJRR、Science Robotics 等期刊进行快速结构化搜索。
- 2. 备用数据源：
-
Google Scholar：通过实时浏览器自动化抓取关键词和特定期刊的数据。
- 3. 筛选条件：
- 关键词：
> 肌肉参数估计、强化学习、模仿学习、人类意图预测、人机交互协作。
- 顶级期刊：TRO、IJRR、RAL、Science Robotics、TMech、CVPR、ICLR、ICCV。
- 日期：按年份筛选，优先关注近期发表（例如 自2024年起）。
- 4. 数据输出：
结构化的 Markdown 报告，保存至 ~/Desktop/YYYY-MM-DD-学术摘要.md，包含以下字段：
- 标题
- 来源期刊（例如 Science Robotics）
- 年份
- 作者
- 摘要
- 主要贡献/创新点
示例配置
yaml
paper_scout SKILL.yaml
crawl:
sources:
- crossref
- scholar_google
queries:
custom: [] # 用户提供
defaults:
- reinforcement learning robotics
- human intention prediction imitation learning
filters:
venues: [TRO, IJRR, Science Robotics, RAL, CVPR]
date: since:2024
output:
path: ~/Desktop/{{today}}-学术摘要.md
markdown: true
示例用法
- 1. 查询 CrossRef 中自 2024 年以来关于 reinforcement learning robotics 的论文。
- 通过抓取 Google Scholar（按日期排序）补充缺失信息。
- 去除重复项，将结果筛选至顶级学术期刊。
- 将 Markdown 摘要保存至桌面。
备注
- - 自动实现跨数据源整合并去除重复项。
- 可扩展至更多专业数据源（例如 IEEE、PubMed）。
- 浏览器抓取（Google Scholar）需要交互式会话以处理 JavaScript 密集型页面。
