book-to-skill vs PageIndex
The two are close in size: 35.4k stars for book-to-skill, 39.3k for PageIndex. Over the days we have tracked them book-to-skill moved +14.8% and PageIndex +15.1%, so PageIndex is growing faster right now.
Neither one leads on the six capability axes, so the choice comes down to which of them fits the way you already work.
Stars and commit dates come from our own daily tracking. The six axes are read off each project's documentation by our review pipeline, so they describe what a project says about itself, not what we measured in its code.
Where they stand today
Converts books and document collections into cross-agent, on-demand 'skills' with per-chapter files so agents answer from real content without dumping entire books into context.
- Stars
- 35.4k
- Tracked growth
- +14.8%
- Maturity
- ●●●●●
- Last commit
- 1d ago
- Language
- Python
- License
- MIT
- Cost to run
- Free, local processing; external model API costs apply
Vectorless, reasoning-based hierarchical retrieval that builds a human-like table-of-contents tree for traceable, context-aware RAG without vector DBs or chunking.
- Stars
- 39.3k
- Tracked growth
- +15.1%
- Maturity
- ●●●●●
- Last commit
- 13m ago
- Language
- Python
- License
- MIT
- Cost to run
- Uses your LLM API key (e.g. OpenAI); cloud service may be paid.
Six axes, head to head
Each axis runs 0 to 5. The label under a score is what that project's own docs claim, not a category average.
| Axis | book-to-skill | PageIndex |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Whole-skill access | ●●●●● Whole-repo analysis |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● Validation & rules | ●●●●● Basic filters/config |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● Config file + options | ●●●●● Config file options |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Local processing | ●●●●● Self-hostable |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● BYO model | ●●●●● Bring-your-own |
Setup ease What it takes to get a first useful run out of it. | ●●●●● One-command install | ●●●●● API key + install |
Which one to pick
Pick book-to-skill if…
Privacy-first — processes documents locally into compact, on-demand agent skills to avoid hallucination and reduce token/token-costs when querying.
Pick PageIndex if…
Vectorless, reasoning-based retrieval — pick PageIndex when you need traceable, explainable, context-aware answers from long professional documents without a vector DB.
What people want from each one
virgiliojr94/book-to-skill
- PDF: `pdftotext -layout` interleaves columns on multi-column documents, and the word-based token estimate hides the 26x size blowup 👍 0
- [bug] Interrupted-run reuse accepts missing or corrupted extracted text 👍 0
- [bug] BeautifulSoup HTML/EPUB extraction inserts newlines inside inline headings and code tokens 👍 0
VectifyAI/PageIndex
Hacker News: Show HN: PageIndex – Vectorless RAG drew 192 points and 128 comments.
Questions people ask
Is book-to-skill better than PageIndex?
Neither one leads on the six capability axes, so the choice comes down to which of them fits the way you already work. book-to-skill is worth picking when privacy-first — processes documents locally into compact, on-demand agent skills to avoid hallucination and reduce token/token-costs when querying.
Which of book-to-skill and PageIndex keeps my code private?
book-to-skill: Local processing (5/5). PageIndex: Self-hostable (4/5).
What does each one cost to run?
book-to-skill: Free, local processing; external model API costs apply. PageIndex: Uses your LLM API key (e.g. OpenAI); cloud service may be paid..
Full profiles: virgiliojr94/book-to-skill and VectifyAI/PageIndex. Everything else in RAG & retrieval.