Skip to content

MCP Prompts

v2.0 ships 5 curated slash-command workflows. These appear automatically in MCP clients that support prompt discovery (e.g. Claude Desktop, Cursor).

Prompt catalogue

Slash-command Required args Optional args Purpose
/rechtsfrage frage rechtsgebiet Answer a German legal question with exact norm citations
/zitation-checken citation — Resolve a citation (e.g. § 433 Abs. 1 BGB) + Stand-Datum
/norm-erklaeren code, norm — Plain-language explanation with cited cross-references
/recherche topic — Multi-step research using research_topic smart tool
/dsgvo-check aktivitaet — Walk through GDPR Art. 5, 6, 7, 9, 13, 14 against a processing activity

Prompt details

/rechtsfrage

Answers a legal question grounded in the corpus. The prompt orchestrates:

  1. A search_laws call to find candidate norms.
  2. Norm text loading via get_norm.
  3. A structured answer with exact § X Abs. Y SGB/BGB/... citations.

Example: /rechtsfrage frage="Wann haftet ein GmbH-Geschäftsführer persönlich?" rechtsgebiet="Gesellschaftsrecht"

/zitation-checken

Resolves a structured citation to a specific norm, verifies it exists in the corpus, and returns the current Stand-Datum from the source metadata.

Example: /zitation-checken citation="§ 433 Abs. 1 S. 1 BGB"

/norm-erklaeren

Loads a norm's full text and its relationship graph, then produces a plain-language explanation with all cross-referenced norms cited.

Example: /norm-erklaeren code="bgb" norm="par:433"

/recherche

Triggers the research_topic smart tool for multi-step corpus research with LLM ranking and synthesis. See research_topic deep dive.

Example: /recherche topic="Datenschutz bei der Videoüberwachung im Einzelhandel"

/dsgvo-check

Walks through GDPR Articles 5, 6, 7, 9, 13, and 14 against a described processing activity. Each article is checked systematically with norm text loaded from the corpus.

Example: /dsgvo-check aktivitaet="Speicherung von Kundendaten für 3 Jahre nach Vertragsende"

Implementation note

All prompts are declared in src/legal_text_mcp_de/prompts/ and registered via register_prompts(app) in server.py. They are stateless — each invocation is independent.

See also