Connect e-Gov Data Portal to Codex

This server connects your AI client to the e-Gov Data Portal, the Japanese government's official open data catalog operated by the Digital Agency (formerly data.go.jp). It is the cross-government entry point for finding where public data lives: datasets published by 23 ministries and agencies, searchable through the portal's standard CKAN API.

Key features:

• Keyword search across all datasets, with optional filters by publishing ministry, tag and file format, plus a choice of sort order. Each hit includes the dataset title, publisher, last-updated date, available file formats, and whether its contents can be read through the API. Every response also carries facets - the top ministries, tags and formats among the hits - so a broad search can be narrowed without guessing.
• Fetch a dataset's full metadata: description, tags, update frequency, and every resource with its resource ID, format and direct download URL (CSV, Excel, PDF, HTML, ...).
• Read the actual rows of resources the portal has loaded into its CKAN datastore, as JSON with column names and types, filtered by exact column values and paged with an offset. Roughly 3% of tabular resources are loaded this way; the dataset detail marks exactly which ones.
• Discover the catalog's real Japanese tag vocabulary by partial match, so a zero-hit search can be recovered instead of abandoned.
• List the publishing ministries and agencies with their dataset counts.

Outside the datastore, this catalog holds metadata (what data exists and where) rather than the data itself - download those files from the resource URLs it returns. Timestamps are ISO 8601 in UTC, not JST. The API is provided free of charge by the Japanese government under the Public Data Terms of Use (v1.0), which permits commercial use with attribution. No API key or account is required.