Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/exegia/corpora-py/llms.txt

Use this file to discover all available pages before exploring further.

Corpora in Context Fabric are Text-Fabric dataset directories stored on your local filesystem. The MCP server loads one or more of these directories at startup from the path(s) you provide via the --corpus flag or the programmatic corpus_manager.load() call. Each dataset is a self-contained folder of annotated .tf feature files that the graph engine indexes at load time.

​
Dataset Format

A Text-Fabric dataset directory is any folder that contains both otext.tf and otype.tf side by side. These two files are mandatory:
  • otext.tf — defines corpus-level metadata (name, version, description, section hierarchy, text formats).
  • otype.tf — maps every node in the graph to its type (e.g. word, verse, chapter, book).
All other .tf files in the directory are feature files that store annotations on nodes (morphology, lemma, gloss, etc.). Organise datasets by category under ~/.exegia/datasets/ for clarity. Context Fabric does not enforce this structure, but it makes multi-corpus setups easy to manage: 
~/.exegia/datasets/
├── bibles/
│   ├── BHSA/           # BHS Hebrew with Annotations
│   │   ├── otext.tf
│   │   ├── otype.tf
│   │   └── *.tf
│   └── GNT/            # Greek New Testament
├── commentaries/
│   └── my-commentary/
└── books/
    └── my-epub-book/

​
Supported Corpus Categories

The BookCategory enum defines the recognised corpus types. You can use these values when tagging or filtering corpora in your application:
CategoryValueDescription
BiblebibleOld/New Testament texts
QuranquranArabic or translated Quran
TanakhtanakhHebrew Bible
CommentarycommentaryRabbinical, patristic, etc.
LexiconlexiconLexical databases (BDB, BDAG)
DictionarydictionaryTheological dictionaries
DevotionaldevotionalDevotional literature
TheologytheologySystematic theology
HistoryhistoryHistorical texts
PhilosophyphilosophyPhilosophical works
FictionfictionLiterary texts
OtherotherCatch-all
from exegia.models.enums import BookCategory

category = BookCategory.COMMENTARY  # "commentary"

​
Loading a Dataset

Pass a dataset path directly to the cf-mcp entrypoint, or load it programmatically with corpus_manager: 
# CLI — stdio mode (for Claude Desktop and other MCP clients)
uv run cf-mcp --corpus ~/.exegia/datasets/bibles/BHSA

# Load multiple corpora at once
uv run cf-mcp \
  --corpus ~/.exegia/datasets/bibles/BHSA --name BHSA \
  --corpus ~/.exegia/datasets/bibles/GNT  --name GNT

# Python — programmatic usage
from exegia.mcp import mcp, corpus_manager

corpus_manager.load("~/.exegia/datasets/bibles/BHSA", name="BHSA")
mcp.run(transport="stdio")

​
Where to Obtain Datasets

Well-known public Text-Fabric datasets you can fetch directly from git:
  • BHSA (BHS Hebrew Bible with Annotations): github.com/ETCBC/bhsa
  • GNT (Greek New Testament): various Text-Fabric repositories on GitHub
  • Custom books: use the EPUB or HTML converters to create your own datasets
See Fetch from Git for how to clone a public repository and locate its dataset directories automatically.

Fetch from Git

Shallow-clone a git repository and locate all Text-Fabric datasets inside it automatically.

Convert EPUB

Turn any EPUB ebook into a queryable Text-Fabric dataset with a full node hierarchy.

Convert HTML

Convert a directory of HTML files into a Text-Fabric dataset with document and element nodes.

Package as .exg

Bundle a Text-Fabric dataset into a single distributable .exg archive with manifest metadata.

Build docs developers (and LLMs) love

Get started for freeTalk to us