According to AlphaMoat, as of 2026-09, 文献精读小工具 has 4,696 downloads and 0 stars — ranked #2,023 of 5,078 AI skills overall, and #192 of 442 in Knowledge.

Installs

657

Downloads

4.7K

Stars

Overall Rank

#2,023

Top 40% overall · Top 44% in Knowledge

Platform	Installs	Downloads	Stars	Comments
ClawHub ↗	29	1.2K	-	1
SkillHub ↗	657	4.7K	-	—

Platforms count differently and update at different times.

What this skill does

It follows an extract-then-summarize pipeline, pulling text with OCR or pdfplumber before the model writes the summary. A run reads only the config and prompt files in the skill directory and executes the bundled scripts, with the research topic and direction set in the prompt file, and logs or exceptions never echo any key. Configured OCR or language-model providers may receive PDF contents and extracted text. Use only endpoints the user has approved, especially for confidential files.

Profile published 2026-10-27

What you can do with it

Generate Chinese close-reading reports for a batch of paper PDFs.
Screen and summarize literature against a defined research direction.

When to use it

When literature volume demands uniformly formatted reading notes.
When a fixed research topic should constrain how summaries are framed.

Example prompts

Generated from the skill’s documentation as examples; not results of actual runs.

Produce Chinese close-reading summaries for these paper PDFs.

Summarize these papers' core findings around my research direction.

Monthly Trend

Downloads4.7K

Stars-

The counting method changed in 2026-09: skills with the same name are now counted separately by author, and the full catalog is covered. Dashed lines show months before the change; they are not compared with later months, and no growth is computed across the change.

Month	Downloads	MoM	Stars
2026-09	4.7K	—	-
2026-08	3.6K	+33.5%	-
2026-07	2.7K	+26.5%	-
2026-06	2.2K	+15.5%	-
2026-05	1.9K	—	-
2026-03	904	—	-

SKILL.md (original)

The author’s original file, shown as plain text without edits or review by AlphaMoat. Links are not clickable; check commands before running them. Fetched 2026-10-23.

3.1 KB · ClawHub v0.0.2

---
name: paper-review-generator
description: 将论文 PDF 批处理为中文精读总结报告的工作流技能。适用于“PaddleOCR 或 pdfplumber 抽取文本 + 大模型总结论文”场景。使用时仅读取 skill 目录下 config.json 与 prompt.md，并运行 scripts 中所需脚本，用户可在prompt.md中定义用户研究主题与研究方向。
---

# paper-review-generator

## 1) 配置文件约束
- 仅使用当前 skill 目录下两个文件：
  - `config.json`：包含运行所需全部信息（是否 OCR、OCR 参数、总结模型 base_url/model/api_key、可见窗口开关、线程数）

Show all (48 lines)

  - `prompt.md`：总结提示词模板
- `api_key` 支持两种写法：
  - 在config.json中指定
  - 环境变量引用：`${ENV_VAR}`，脚本会在运行时读取对应环境变量
- 不读取其他目录 secret。
- 不在日志和异常信息中回显任何 api_key 或 token。

## 2) 执行入口
- **必须先切换到本 skill 根目录再运行脚本**（即 `.../paper-review-generator`），否则相对路径的 `config.json` / `prompt.md` 会找不到。
- 必须由用户明确传入输入与输出路径：
  - `--pdf`（可重复，支持多个文件）
  - `--dir`（可重复，支持多个文件夹）
  - `--output-dir`（可选；不传时默认输出到每个输入 PDF 同目录下的 `总结` 文件夹）
- 示例（路径由 agent 按用户需求填入）：
  - 单文件：`python scripts/run_pipeline.py --pdf "{pdf_path}" --output-dir "{output_dir}"`
  - 多文件：`python scripts/run_pipeline.py --pdf "{pdf_path_1}" --pdf "{pdf_path_2}" --output-dir "{output_dir}"`
  - 单文件夹：`python scripts/run_pipeline.py --dir "{pdf_dir}" --output-dir "{output_dir}"`
  - 多文件夹：`python scripts/run_pipeline.py --dir "{pdf_dir_1}" --dir "{pdf_dir_2}" --output-dir "{output_dir}"`

## 3) 分流逻辑
- 读取 `config.json.use_paddleocr`：
  - `true`：调用 `extract_paddleocr.py` 抽取文本（JSON 行输出，不落盘）。
  - `false`：调用 `extract_pdfplumber.py` 抽取文本（JSON 行输出，不落盘）。
- 然后调用 `summarize_reports.py`：读取 `prompt.md` 与管道传递的抽取文本，调用 `summarizer.provider` 指定的模型配置生成 `*_研读报告.md`。

## 4) 环境检查与安全规范
- 执行前先检查 Python 是否可用（建议 3.10+）：
  - 若用户电脑未安装 Python，必须先明确提示用户安装 Python，再继续后续步骤。
- 执行前检查依赖：
  - 若缺少依赖包，agent 应在 skill 根目录按 `scripts/requirements.txt` 执行安装：
    - `pip install -r scripts/requirements.txt`
- 首次使用前必须做端点审查：
  - 只保留你信任的 provider，删除或留空不用的 `base_url`/`model`
  - 敏感文档场景优先使用自建/内网 OCR 与 LLM 端点
- 仅向用户明确确认过的 OCR/LLM 端点发请求。
- 若配置缺失（如 api_key/token/model/base_url），直接报错并提示补齐字段。
- 日志与异常必须脱敏，禁止输出原始 Authorization/API key/token 或完整远端响应体。