https://github.com/malloydata/malloyyo-auto-recalls
Malloy semantic model over NHTSA vehicle-recall data
https://github.com/malloydata/malloyyo-auto-recalls
Last synced: 15 days ago
JSON representation
Malloy semantic model over NHTSA vehicle-recall data
- Host: GitHub
- URL: https://github.com/malloydata/malloyyo-auto-recalls
- Owner: malloydata
- Created: 2026-07-02T23:55:07.000Z (23 days ago)
- Default Branch: main
- Last Pushed: 2026-07-04T22:10:47.000Z (21 days ago)
- Last Synced: 2026-07-05T00:07:29.929Z (21 days ago)
- Language: TypeScript
- Size: 4.98 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# malloyyo-auto-recalls
A [Malloy](https://www.malloydata.dev/) semantic model over NHTSA vehicle-recall
data, analyzed by component category and recall trends over time — with three
**model-declared dashboards** (see `docs/repo-artifacts.md` in the malloyyo repo
for the design).
## Contents
- `auto_recalls.malloy` — the `auto_recalls` source: recalls with a derived
`component_category` classification, a lowercase `search_text` haystack,
measures (`recall_count`, `total_affected_vehicles`, …), the dashboard givens,
a `manufacturer_options` suggestions query, and three `# artifact`-tagged
dashboard queries.
- `index.malloy` — entry point that re-exports the source, givens, and queries.
- `dashboards//Dashboard.tsx` — the dashboards' custom React components.
Data source: `https://storage.googleapis.com/malloyyo/auto_recalls/auto_recalls.parquet`
(NHTSA recalls: https://www.nhtsa.gov/recalls)
## Dashboards — declared in the model, not in JSON
There is no manifest file. Each dashboard is a query tagged `# artifact`; the
filters are `filter` givens applied with `~`, so **an empty expression means
"no filter"** — `recall_list` needs no `($X = '' or …)` sentinel logic:
```malloy
given:
# label="Manufacturer" control=select suggest { query=manufacturer_options dimension=Manufacturer }
MANUFACTURER :: filter is f''
# label="Component" control=select suggest { source=auto_recalls dimension=Component }
COMPONENT :: filter is f''
# label="Park outside advisory"
PARK_OUTSIDE :: boolean is false
# artifact { name="manufacturer" title="Manufacturer Recall Profile" givens { MANUFACTURER="Ford Motor Company" } }
query: manufacturer_dashboard is auto_recalls -> { where: Manufacturer ~ $MANUFACTURER … }
```
Notable patterns exercised here:
- **Curated options via a named query** — `suggest { query=manufacturer_options
dimension=Manufacturer }` fills the manufacturer dropdown from the top-60-by-
recall-count query (governed, reviewable Malloy — not a string in a tag).
- **Distinct-values options** — `suggest { source=auto_recalls dimension=Component }`.
- **Per-dashboard defaults** — the profile starts on Ford, the recall list on
"everything", from the same shared given (`# artifact { … givens { … } }`).
- **Search** — the `SEARCH` given filters the `search_text` dimension; the
components commit `filters.contains(term)` so escaping is always correct
(values like `Tesla, Inc.` must be committed with `filters.oneOf`, never raw —
a bare comma parses as filter alternatives).
- **Boolean givens** → the `` widget.
| slug | query | notes |
|---|---|---|
| `manufacturer` | `manufacturer_dashboard` | select fed by `manufacturer_options`, Ford default |
| `vehicle-search` | `vehicle_search_dashboard` | debounced search, `%tundra%` default |
| `recall-list` | `recall_list` | every filter optional, Checkbox widgets, Clear button |
## Running
- `malloyyo dashboard dev` — preview server with hot reload.
- `malloyyo lint` — validates the tagged queries, given suggest declarations,
and the `Dashboard.tsx` files.
- Or open the folder in VS Code with the
[Malloy extension](https://marketplace.visualstudio.com/items?itemName=malloydata.malloy-vscode),
or use the Malloyyo MCP server configured in `.mcp.json`.