Ielts Extractor

自动从剑桥雅思PDF中提取多页连续阅读文章和题目内容，支持双栏排版并保存为结构化JSON格式。

Security Scan

The skill's description promises full extraction of passages and questions into structured JSON, but the provided code and instructions are inconsistent and incomplete (missing question parsing, JSON write, and declared dependencies).

Purpose & Capability

The name/description promise: extract multi-page Cambridge IELTS passages and questions, handle two-column layout, and save structured JSON. The included Python (extract_article.py) only locates a passage and concatenates page text (no question parsing, no JSON output, and the two-column helper is defined but not used). References describe image extraction with PyMuPDF and JSON schema, but those behaviors are not implemented in the code.

Instruction Scope

SKILL.md describes locating passages, extracting multi-page text, handling two-column layout, extracting question groups and options, and saving JSON to specific ielts-tracker paths. The runtime instructions reference writing files into project directories (ielts-tracker/.../public/images/) and expect question extraction, but the actual code does not perform those writes or parse questions. Instructions don't request unrelated credentials or environment variables.

Install Mechanism

No install spec (instruction-only) — lower risk — but the code imports pdfplumber and references fitz/PyMuPDF in docs. Dependencies are not declared; runtime may fail or require installing third-party packages. No external downloads or suspicious URLs are present.

Credentials

The skill requests no environment variables, no credentials, and no config paths. However, SKILL.md and references expect writing output files into project paths which could overwrite local files if run with file-system access; this is expected for a data-extraction tool but worth noting.

Persistence & Privilege

always is false and the skill is user-invocable. It does not request permanent agent inclusion or modify other skills. There is no autonomously elevated privilege requested.

What to consider before installing

This skill is internally inconsistent rather than obviously malicious: its description and docs promise full passage+question extraction and JSON output, but the shipped code only extracts concatenated article text and prints it. Before using, ask the author to (1) implement or remove question-parsing and JSON write logic, (2) declare/install dependencies (pdfplumber, and optionally PyMuPDF/fitz), and (3) confirm where files will be written to avoid overwriting local project files. Test the script on non-sensitive sample PDFs in a sandbox or disposable directory to verify behavior. If you don't trust the author, do not run the code on sensitive systems or give it write access to important directories.

IELTS 试题数据提取 Skill

概述

从剑桥雅思 PDF 提取阅读文章和题目。

触发条件

用户要求"提取雅思试题"时使用。

流程(4步)

1. 定位 PDF 和页码

• 查找 "Test X" + "READING PASSAGE Y"
• 记录起始页码

2. 提取文章

• 必须连续提取多页
• 处理两栏布局
• 检查字数: 1500-2500词/篇

详见: references/pdf-extraction.md

3. 提取题目

• 按大题分组
• 选择正确题型
• 完整选项(单选每题A-D)

详见: references/question-types.md

4. 保存 JSON

• 使用 content 字段
• 选项格式正确

详见: references/json-format.md

数据文件

ielts-tracker/data/tests/cambridge-{4,5,6}/test-{1-4}/test.json
ielts-tracker/ielts-app/public/images/

题型速查

题型	type
标题配对	matching-headings
判断题	yes-no-not-given
单选	multiple-choice-single
多选	multiple-answer
表格	table-completion
填空	fill-blank-summary