bht-mcp 0.1.0

MCP server for the BHt (Biblia Hebraica transcripta) Hebrew Bible database

bht-mcp

MCP server for the BHt (Biblia Hebraica transcripta) Hebrew Bible database at LMU Munich.

Gives AI assistants direct access to 489,000+ morphologically analyzed Hebrew Bible tokens — search by book, root, part of speech, or any of 42 linguistic fields.

한국어 안내는 아래에 있습니다.

Quick Start

1. Verify Python

python3 --version  # 3.11+

2. Install

pip install bht-mcp

3. Configure your MCP client

Add to your client's MCP server configuration:

{
  "mcpServers": {
    "bht": {
      "command": "bht-mcp"
    }
  }
}

That's it. The server runs locally via stdio — no API keys, no remote server, no cost.

What You Can Do

Verse Analysis

"Show me Genesis 1:1 word by word"

bht_search(filters=[{field:"buch", value:"Gen"}, {field:"kapitel", value:"1"}, {field:"vers", value:"1"}])
→ 11 tokens: b˙, rē(ʾ)šīt, barā(ʾ), ʾïlō*hīm, ʾȧt, ha, šamaym, w˙, ʾȧt, ha, ʾarṣ

bht_token_detail(buch="Gen", beleg_nr=3)
→ barā(ʾ): Suffixkonjugation, 3ms, Qal, root BRʾ, meaning "fac,prod"

Vocabulary Study

"Find all occurrences of root BRʾ (create)"

bht_search(filters=[{field:"Wurzel", value:"BRʾ"}])
→ All tokens derived from this root across the entire Hebrew Bible

Grammar Patterns

"Show all Qal verbs in Psalms"

bht_search(filters=[{field:"buch", value:"Ps"}, {field:"wa", value:"11 VERB"}, {field:"stamm", value:"G"}])

Syntax

"What's the syntactic structure of Genesis 1:1?"

bht_syntax_tree(buch="Gen", kapitel=1, vers=1, satz="PR")
→ KOORDV → [PV → [PRAEP, ATKV → [ATK, SUB]], KONJS → [KONJ, PV → [...]]]

bht_sentence_analysis(buch="Gen", kapitel=1, vers=1, satz="PR")
→ Satzart: V4.1, Syntagmen: P(0 1) 1(1 2) 2(2 9)

Textual Criticism

"Are there manuscript variants in Genesis 1?"

bht_text_annotations(buch="Gen", kapitel=1)
→ 4 annotations including Greek/MT variants

Tools

Tool	Purpose	BHt Requests
bht_list_books	List all 47 books with codes and chapter counts	0
bht_field_info	Get valid values for any of 42 search fields	0–1
bht_search	Search tokens by location or morphological filters	0–1
bht_token_detail	Full morphological analysis of a single token	0–1
bht_syntax_tree	Word-level syntactic tree (Wortfügungsebene)	0–2
bht_sentence_analysis	Sentence-level analysis (Satzfügungsebene)	0–2
bht_text_annotations	Textual criticism annotations for a chapter	0–1

Book Codes

Torah:    Gen Ex Lev Num Dt
Prophets: Jos Ri 1Sam 2Sam 1Koen 2Koen Jes Jer Ez
          Hos Joel Am Ob Jon Mich Nah Hab Zef Hag Sach Mal
Writings: Ps Ij Spr Rut Hl Koh Klgl Est Dan Esr Neh 1Chr 2Chr
Sirach:   ASir BSir CSir DSir ESir MSir QSir TSir

Common Search Fields

Field	Description	Example Values
buch	Book code	Gen, Ps, Jes
wa	Part of speech	11 VERB, 12 SUBSTANTIV, 31 PRAEPOSITION
stamm	Verbal stem	G(Qal), D(Piel), H(Hiphil), N(Niphal)
ps	Person	1, 2, 3, 0(n/a)
gen	Gender	M, F, 0(n/a)
num	Number	S(sg), P(pl), D(dual), 0(n/a)
Wurzel	Root	BRʾ, ʾMR, HLK
lexem	Lexeme	baraʾ l+, ʾilōhīm l+
basis	Base form	brʾ 1, ʾl-h

Use bht_field_info(field="wa") to get the full list of valid values for any field.

How It Works

┌──────────────┐  stdio  ┌──────────────────┐  cache miss  ┌─────────────┐
│  MCP Client  │ ───────→│  bht-mcp         │ ───────────→│  BHt Website │
│  (your AI)   │ ←───────│  (local process) │ ←───────────│  (LMU Munich)│
└──────────────┘         │  ┌────────────┐  │              └─────────────┘
                         │  │ ~/.bht/    │  │
                         │  │ cache.db   │  │  ← local SQLite cache
                         │  └────────────┘  │
                         └──────────────────┘

Progressive caching: The first search for a book fetches all its tokens (one API call). Subsequent searches for the same book use the local cache — zero network requests. Morphological detail is cached per-token as you explore.

Rate Limits

This tool accesses a university research server. Built-in limits protect it:

Limit	Value	Reason
Request interval	1 req/s	Server capacity protection
Daily HTML requests	150/day	Beleg, tree, sentence, annotation pages
Daily JSON requests	Unlimited	Search API is lightweight

These limits are comparable to a researcher manually browsing the site (~25–100 pages/day). The 150/day limit resets at midnight.

Every tool response includes a quota field showing current usage:

{
  "data": [...],
  "quota": {"daily_html_used": 12, "daily_html_limit": 150, "daily_html_remaining": 138}
}

About BHt

BHt (Biblia Hebraica transcripta) is a digital transcription of the Hebrew Bible maintained at LMU Munich under the direction of Prof. Wolfgang Richter. It features:

• 489,437 tokens with full morphological analysis
• Richter transcription system (Latin-script phonemic representation)
• 5-class part-of-speech system (Hauptwortart, Nebenwortart, Fügwortart, etc.)
• Syntactic trees at word-level (Wortfügungsebene) and sentence-level (Satzfügungsebene)
• Textual criticism annotations with manuscript variant data

The transcription uses a distinctive system where e.g. barā(ʾ) represents the Hebrew word בָּרָא, with morphophonemic detail not found in other digital Hebrew Bible projects.

License

Code: MIT

BHt data: CC BY-SA 4.0 (LMU Munich)

---

한국어

설치

pip install bht-mcp

MCP 클라이언트 설정

{
  "mcpServers": {
    "bht": {
      "command": "bht-mcp"
    }
  }
}

사용 예시

구절 분석: "창세기 1:1을 단어별로 보여주세요"

• bht_search로 토큰 검색 → bht_token_detail로 형태론 상세 확인

어휘 연구: "어근 BRʾ(창조하다)의 모든 용례를 찾아주세요"

• bht_search(filters=[{field:"Wurzel", value:"BRʾ"}])

문법 패턴: "시편의 모든 칼(Qal) 동사를 찾아주세요"

• bht_search(filters=[{field:"buch", value:"Ps"}, {field:"wa", value:"11 VERB"}, {field:"stamm", value:"G"}])

통사론: "창세기 1:1의 문장 구조는?"

• bht_syntax_tree로 단어 수준 통사 트리, bht_sentence_analysis로 문장 수준 분석

텍스트 비평: "창세기 1장에 사본 이본이 있나요?"

• bht_text_annotations(buch="Gen", kapitel=1) → 그리스어/마소라 사본 차이

작동 방식

로컬에서 실행됩니다. 원격 서버 없음, API 키 없음, 비용 없음.

• 첫 검색 시 해당 책의 전체 토큰을 1회 API 호출로 가져와 로컬 SQLite에 캐시
• 이후 같은 책의 검색은 네트워크 요청 0회
• 형태론 상세(beleg)는 토큰별로 on-demand 캐시

제한사항

대학 연구 서버를 보호하기 위한 내장 제한:

• 초당 1요청 (서버 부하 보호)
• 일일 HTML 요청 150건 (beleg, tree, sentence, annotation)
• 일일 JSON 요청 무제한 (검색 API는 경량)

라이선스

코드: MIT | BHt 데이터: CC BY-SA 4.0 (LMU Munich)