Harvest-API: `POST /api/validate` with `fetch_method=url` returns 500 with HTML page
- Dominant language
- Python
- Stars
- 1.1k
- Forks
- 200
- Avg merge
- 4h 56m
- Merged PRs (30d)
- 5
Description
**Description**
The /api/validate endpoint returns HTTP 500 with an HTML error page when fetch_method=url is used. The paste method works correctly.
**Steps to Reproduce**
POST [https://harvest.data.gov/api/validate Content-Type](https://harvest.data.gov/api/validateContent-Type): application/json
{ "fetch_method": "url", "schema": "dcatus1.1: federal dataset", "url": "https://www.gsa.gov/data.json" }
**Expected Behavior**
JSON response containing validation results.
**Actual Behavior**
HTTP 500 with an HTML "Site unavailable" page.
**Notes**
Confirmed April 27, 2026. The paste method (fetch_method=url replaced with fetch_method=paste and json_text provided) works correctly and returns {"validation_errors": [...]}. Only the URL fetch method is broken. This is the primary intended use case for agencies validating their data.json before harvest -- high priority fix. Needs testing with additional agency URLs to rule out timeout issues specific to [gsa.gov](http://gsa.gov/).
Contributor guide
Assessment
This issue has not been assessed yet.