\r \r

Book Summarizer\r

\r Use this skill for requests like:\r \r

• "Summarize this book at a 20% compression ratio"\r
• "Generate a substantial summary and verify the ratio"\r
• "Produce the summary in batches and validate the final total"\r \r

Rules\r

\r

• Default target ratio: 0.20\r
• Default tolerance: 0.02\r
• Accept only summaries between 18% and 22% of the original word count\r
• Output must be in user's language of choice (e.g., pt-BR)\r
• If the ratio is outside the allowed range, do not import the summary\r
• In chat-only mode, if the source is very large, generate the summary in multiple batches and merge them before validation\r
• All packaged helper scripts live in scripts/ inside this skill folder\r \r

Workflow\r

\r

1. Start from a local plain-text book file or a downloaded Project Gutenberg text.\r
2. Count the source words with python scripts/book_tools.py count \x3Coriginal_file>.\r
3. Compute the target summary length with python scripts/book_tools.py target \x3Coriginal_file> --ratio 0.20.\r
4. If the source is too large for one reply, split it with python scripts/split_book.py \x3Coriginal_file> 3000.\r
5. Draft summary batches in order, preserving chronology and section fidelity.\r
6. Merge the batches with python scripts/book_tools.py aggregate \x3Csummary_file> \x3Cbatch_files...>.\r
7. Validate the final ratio with python scripts/verify_summary_ratio.py \x3Coriginal_file> \x3Csummary_file>.\r \r

Key Files\r

\r

• scripts/book_tools.py\r
• scripts/split_book.py\r
• scripts/verify_summary_ratio.py\r
• SKILL.md\r \r

Notes\r

\r

• The packaged scripts use only the Python standard library.\r
• Run the commands from the skill folder, or use explicit paths if you call them from elsewhere.\r
• For very large books, prefer the automated pipeline over single-turn chat drafting.\r
• In chat-only mode, books above roughly 80k words should be summarized over multiple turns; do not pretend a single short draft satisfies the 20% rule.\r \r \r