When to use this skill
When the user asks to download a full Statistics Canada data table — population estimates, GDP, labour force, trade data — or says "get me the CSV for table 14-10-0287" or "download StatCan data." This skill gets you a download URL for a complete table file. For querying specific time-series values without bulk download, use get-statcan-time-series instead. For browsing which tables exist, use browse-statcan-table-catalog.
Your best first call
curl "https://www150.statcan.gc.ca/t1/wds/getFullTableDownloadCSV/14100287/en"
No auth. No key. Returns a JSON envelope where object[0].downloadLink is the URL for the full CSV file (often ZIP-compressed). Replace 14100287 with the 10-digit product ID you need; switch en to fr for French.
The response shape:
status — "SUCCESS" or an error codeobject[0].downloadLink — the URL to fetch the actual data file fromobject[0].productId — confirms which product was requestedobject[0].releaseTime — timestamp of the most recent data update for this table
The endpoint does not return the data itself — it returns a URL you must then fetch separately to get the CSV.
Fallbacks (when the best call isn't enough)
- Don't know the product ID →
getAllCubesListLite returns the full table catalog (~600 entries) with productId, cansimId, cubeTitleEn, and frequencyCode. Search it to find the right productId before calling getFullTableDownloadCSV.
- Need SDMX instead of CSV →
getFullTableDownloadSDMX/{pid} returns the same envelope with an SDMX download link. Use only when the user explicitly requests SDMX or needs SDMX tooling integration.
- Need a specific value, not the whole table →
getDataFromVectorByReferencePeriodRange fetches individual data points by vector ID and date range. Use when the user wants one number or a short series, not a bulk download.
Pitfalls
getFullTableDownloadCSV returns a download URL, not the data itself. It is always two calls: one to get the link, one to fetch the CSV from that link.- Product IDs are opaque 10-digit numbers (e.g.
14100287 for population estimates). If the user says "population," you must look up the productId via getAllCubesListLite first — there is no text search endpoint.
- The
language path segment accepts only en or fr. Any other value returns an error.
getAllCubesListLite returns the entire catalog in one unparameterized call. Fetch it once per session and index by cubeTitleEn rather than calling it repeatedly.
One-line summary for the user
I can get a download link for any Statistics Canada data table — population, GDP, trade, labour — by product ID, no auth required, then fetch the CSV.
REST API for accessing Statistics Canada data including economic indicators, census data, and statistical tables. Supports querying by product ID, vector ID, and coordinates.
SKILL.md source (frontmatter + body)
---
name: access-ZIP
description: When the user asks to download a Statistics Canada data table — GDP, population, labour force, trade — or needs a CSV/ZIP file for a known product ID. Statistics Canada WDS returns the download URL in one unauthenticated GET.
---
## When to use this skill
When the user asks to download a full Statistics Canada data table — population estimates, GDP, labour force, trade data — or says "get me the CSV for table 14-10-0287" or "download StatCan data." This skill gets you a download URL for a complete table file. For querying specific time-series values without bulk download, use `get-statcan-time-series` instead. For browsing which tables exist, use `browse-statcan-table-catalog`.
## Your best first call
```bash
curl "https://www150.statcan.gc.ca/t1/wds/getFullTableDownloadCSV/14100287/en"
```
No auth. No key. Returns a JSON envelope where `object[0].downloadLink` is the URL for the full CSV file (often ZIP-compressed). Replace `14100287` with the 10-digit product ID you need; switch `en` to `fr` for French.
The response shape:
- `status` — `"SUCCESS"` or an error code
- `object[0].downloadLink` — the URL to fetch the actual data file from
- `object[0].productId` — confirms which product was requested
- `object[0].releaseTime` — timestamp of the most recent data update for this table
The endpoint does not return the data itself — it returns a URL you must then fetch separately to get the CSV.
## Fallbacks (when the best call isn't enough)
- **Don't know the product ID** → `getAllCubesListLite` returns the full table catalog (~600 entries) with `productId`, `cansimId`, `cubeTitleEn`, and `frequencyCode`. Search it to find the right `productId` before calling `getFullTableDownloadCSV`.
- **Need SDMX instead of CSV** → `getFullTableDownloadSDMX/{pid}` returns the same envelope with an SDMX download link. Use only when the user explicitly requests SDMX or needs SDMX tooling integration.
- **Need a specific value, not the whole table** → `getDataFromVectorByReferencePeriodRange` fetches individual data points by vector ID and date range. Use when the user wants one number or a short series, not a bulk download.
## Pitfalls
- `getFullTableDownloadCSV` returns a download URL, not the data itself. It is always two calls: one to get the link, one to fetch the CSV from that link.
- Product IDs are opaque 10-digit numbers (e.g. `14100287` for population estimates). If the user says "population," you must look up the `productId` via `getAllCubesListLite` first — there is no text search endpoint.
- The `language` path segment accepts only `en` or `fr`. Any other value returns an error.
- `getAllCubesListLite` returns the entire catalog in one unparameterized call. Fetch it once per session and index by `cubeTitleEn` rather than calling it repeatedly.
## One-line summary for the user
I can get a download link for any Statistics Canada data table — population, GDP, trade, labour — by product ID, no auth required, then fetch the CSV.