Skip to content

get_source_limitations

Return official-source limitations with optional metadata filters.

This tool exposes the non-imported portion of the corpus — sources that exist in the manifest but could not be fully ingested. Downstream consumers can use this to understand coverage gaps before relying on the corpus for a specific law.

Parameters

Name Type Required Description
source_family string no Filter by source family: gesetze-im-internet or eur-lex.
terminal_state string no Filter by terminal state: parse_failed, source_unavailable, unsupported_format, excluded_by_policy.
state_code string no Filter by German state code (for Länder laws).
law_id string no Filter by a specific law identifier.

Returns

{
  limitations: Array<{
    law_id: string;
    source_family: string;
    terminal_state: string;
    reason: string;
    source_url: string | null;
    state_code: string | null;
  }>;
  total: number;
}

Example

All limitations:

result = mcp_client.call_tool("get_source_limitations", {})

Parse failures only:

result = mcp_client.call_tool("get_source_limitations", {
    "terminal_state": "parse_failed"
})

{
  "limitations": [
    {
      "law_id": "somelaw",
      "source_family": "gesetze-im-internet",
      "terminal_state": "parse_failed",
      "reason": "XML structure not recognized by parser",
      "source_url": "https://www.gesetze-im-internet.de/somelaw/",
      "state_code": null
    }
  ],
  "total": 1
}

Notes

  • Returns an empty limitations array when all sources are imported.
  • Each non-imported source has exactly one terminal state; there are no partial states.
  • Use get_corpus_coverage to get aggregate counts by terminal state.

Source

mcp/server.py — @app.tool() definition for get_source_limitations.