name: Book Ingestion Skill description: Instructions for extracting content from raw book text into the application's data structure.

Book Ingestion Skill

This skill provides a standard operating procedure for converting raw text from vocabulary books (Grade 4-10) into the application's TypeScript data format.

Relevant Files

• Input Example: skills/book_ingestion/examples/input.txt
• Output Example: skills/book_ingestion/examples/output.ts
• Target Data Schema: lib/types.ts

Process Overview

1. Acquire Content:
   • Text Files: Look for raw text files in text/ (e.g., text/book_4_lesson_20.txt).
   • Images: If only images exist (e.g., text/-106_1.jpg), use the view_file tool to "read" the image. You must transcribe the text from the image into the Raw Text patterns below before or during the conversion process.
2. Analyze Structure: Identify the four main components: Word List, Exercises, Reading Passage, and Fun & Fascinating FACTS.
3. Convert to TypeScript: Use the patterns below to generate the data.ts file.

Text Extraction Patterns

1. Vocabulary List

Input Pattern:

word
(pos.) Definition.
(pos2.) Definition 2.
Example sentence.

Output Logic:

• ID: lowercase word.
• Word: original word.
• Definition: Concatenate all parts of speech and definitions.
• Example Sentence: The final line.
• Emoji: Pick a relevant emoji manually or using a tool.
• Category: "Grade [X]"

2. Exercises (Stories)

Exercises are treated as Story objects with type: "exercise".

Common Types:

• Finding Meanings: Choose two phrases (e.g., "c-a").
• Just the Right Word: Replace bold phrase with a word.
• Applying Meanings: Circle correct answers (multiple choice).
• Word Study: Prefix/Suffix analysis, Synonyms, or Analogies. Ingest this section.
• EXCLUDE: none.

Output Logic:

• IDs: b[Book]-l[Lesson]-s[Index] (e.g., b4-l19-s1).
• Title: Exercise name (e.g., "19A Finding Meanings").
• Content: The instructions text.
• Questions: Array of Question objects.
  • ID: [exercise-id]-q[Number].
  • Type: usually "multiple-choice".
  • Options: The provided choices.

3. Reading Passage

Input Pattern:

Reading Passage
Title
[Paragraphs...]

Output Logic:

• ID: b[Book]-l[Lesson]-s[Last]
• Title: "Reading Passage: [Title]"
• Type: "non-fiction"
• Content: The full text. Important: Bold specific vocabulary words using markdown **word**.
• Questions: Array of Question objects.
  • ID: b[Book]-l[Lesson]-p-q[Number]
  • Type: "multiple-choice"
  • Conversion Requirement: You must convert the original 15 open-ended questions into multiple-choice questions.
  • Structure:
    • question: The original question text.
    • options: 4 distinct options including the correct answer.
    • correctAnswer: The correct answer string.

4. Fun & Fascinating Facts

Input Pattern:

Fun & Fascinating FACTS
[Points...]

Output Logic:

• ID: b[Book]-l[Lesson]-s[Last+1] (e.g., b6-l1-facts).
• Title: "Fun & Fascinating FACTS".
• Type: "non-fiction".
• Content: The full text, usually bulleted points. Bold vocabulary words.
• wordsIncluded: List of words mentioned/taught in the facts.

Validation

After generating the file:

1. Check strictly against lib/types.ts.
2. Verify array exports match the naming convention: b[X]_l[Y]_words, b[X]_l[Y]_story_[Z].