librarian
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LIB_HOME | No | Path to the library root directory (default ./library) | ./library |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| list_booksA | List of all books in the library: id, author, title, chapter count, token count, quality status. |
| list_chaptersB | Table of contents for a book by its id: chapter number, title, token count, and a short summary for each chapter. |
| findA | Full-text search across chapters and book titles/authors in the whole library (bm25 ranking, snippets, RU/EN stemming). The book_id parameter restricts the search to one book. |
| get_chaptersA | Chapter text of a book: by chapter numbers (spec, e.g. "1-3,5") or by token budget (budget). If neither spec nor budget is given, the default budget of 12000 tokens is used. Passing both parameters at once is an error. If even the first chapter doesn't fit the budget, an empty text with an explanation in message is returned instead of an error — retry with a larger budget or use spec. |
| book_infoC | Metadata and quality metrics for a book: book, metrics, subscores, score, hard_triggers — the same projection as |
| verify_quoteA | Verify a quote against the library: an exact/close/distorted/not_found verdict, the book/chapter location, similarity, a passage of surrounding text, and a word-level diff against the source. Provide book_id to check one specific book (full scan, no FTS5 needed); omit it to search the whole library via FTS5 candidates. Use this whenever you quote a book to the user. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 6 tools
Each tool has a clearly distinct purpose: listing books, listing chapters, full-text search, retrieving chapter text, detailed metadata, and quote verification. The overlap between list_books and book_info is minimal and well-separated by granularity.
Naming is a mix of patterns: list_books/list_chapters use a consistent list_ prefix, get_chapters follows get_ style, while find, book_info, and verify_quote deviate. The inconsistency is readable but not uniform.
Six tools is an appropriate, focused set for a librarian server—neither too sparse nor bloated. Each tool covers a distinct access/verification need without redundancy.
The tool surface covers the core read-only library operations: browsing books, viewing TOCs, reading chapter text, searching, getting detailed metadata, and verifying quotes. There are no obvious missing features for this domain.