Open data finder

Locate the right public dataset and preview it with its licence.

Related projects: None in the catalogue.

Calling the tools

Steps name each operation as source/operation (e.g. ppr/ppr_price_stats), using the typed tool name as the operation. On the default surface, run it with ireland_call, for example:

{ "source": "data-gov-ie", "operation": "datagov_search_datasets", "args": { "query": "bike counts" } }

If an argument shape is unclear, call ireland_describe with { "source", "operation" } first; it returns the schema and an example. search, fetch and nearby are top-level tools. If the server was added with ?toolsets=<ids> or ?toolsets=all, the typed tool (e.g. datagov_search_datasets) can be called directly with the same args.

Steps

1. Search broadly. Call search with the topic to see hits across all sources, then call fetch with a promising id (for example data-gov-ie:moby-bikes).
2. Search the portals. Call data-gov-ie/datagov_search_datasets (national) and, for Dublin topics, smart-dublin/smartdublin_search_datasets with the topic words.
3. Inspect. Call data-gov-ie/datagov_get_dataset (or smart-dublin/smartdublin_get_dataset) for the chosen dataset: publisher, licence, last modified date and resources (CSV, API, etc.).
4. Preview rows. If a resource is in the datastore, call data-gov-ie/datagov_query_datastore (or smart-dublin/smartdublin_query_datastore) with the resource id and a small limit.
5. If a dedicated source covers the topic (CSO, PPR, Met Éireann, OPW, etc.), suggest that source's ops instead, because they are cleaner than raw portal files. cross/list_sources lists them.

Output format

| Dataset | Publisher | Licence | Updated | Formats | Link |
|---|---|---|---|---|---|
Preview (<resource>, first 5 rows): <table>
Sources
- data.gov.ie catalogue; data © <publisher> (<licence>)

Rules

• Follow ../_shared/citations.md. Licences are per dataset: report each one.
• Show the dataset's last-modified date; flag datasets not updated in over 2 years as possibly stale.
• Do not claim a dataset is complete or authoritative beyond what its publisher states.