`get_source_metadata`¶

Return source provenance for all laws or one law code/alias.

Parameters¶

Name	Type	Required	Description
`code`	`string`	no	Law abbreviation or alias. If omitted, returns provenance for all loaded laws.

Returns¶

When code is provided:

{
  code: string;
  source_url: string;       // original official source URL
  fetch_timestamp: string;  // ISO 8601 timestamp of the fetch
  content_hash: string;     // SHA-256 of the source payload
  terminal_state: string;   // "imported" for successfully loaded laws
  source_family: string;    // "gesetze-im-internet" or "eur-lex"
}

When code is omitted, returns an object with a sources array of the above shape.

Example¶

Provenance for a single law:

result = mcp_client.call_tool("get_source_metadata", {"code": "bgb"})

{
  "code": "bgb",
  "source_url": "https://www.gesetze-im-internet.de/bgb/",
  "fetch_timestamp": "2026-05-15T10:23:00Z",
  "content_hash": "sha256:abc123...",
  "terminal_state": "imported",
  "source_family": "gesetze-im-internet"
}

All sources:

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

Notes¶

This tool returns only successfully imported laws. For laws that failed to load, use get_source_limitations.
The content_hash can be used to detect if the upstream source has changed since the corpus was generated.

get_source_limitations — limitations and non-imported sources.
get_corpus_coverage — aggregate coverage statistics.

Source¶

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

get_source_metadata¶