[Itstommy10]

Summary

Cursor's usage endpoint returns some spendLimitUsage fields as strings instead of numbers (seen on a Free account: "pooledLimit":"0"). Because CursorSpendLimitUsage declares them as int, decoding the whole usage response failed and Cursor polling never stored data:

Failed to fetch Cursor quotas error="cursor: fetch usage: json: cannot unmarshal string into Go struct field CursorSpendLimitUsage.spendLimitUsage.pooledLimit of type int"

This adds a custom UnmarshalJSON for CursorSpendLimitUsage that accepts the integer fields as JSON numbers, numeric strings, empty strings or null (empty/null/non-numeric decode as 0). The struct fields and all call sites stay int, so nothing else changes.

Test plan

• New tests: numbers, numeric strings, empty/null, float, nested in the usage response
• go vet ./internal/api/ and go test ./internal/api/
• Verified against a real Free account: polling now completes (Cursor poll complete account_type=individual plan_name=Free quota_count=2)

[codecov]

Codecov Report

❌ Patch coverage is 74.35897% with 10 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
internal/api/cursor_types.go	74.35%	5 Missing and 5 partials ⚠️

📢 Thoughts on this report? Let us know!

[prakersh]

Thanks a lot for digging into this and sending a fix, @Itstommy10, and for checking it against a real Free account. That confirmation is really useful.

The same issue was reported as #151, and the fix landed on main in #153 and ships in v2.14.14. It adds string-tolerant decoders for every numeric field in the Cursor responses, not just spendLimitUsage, and also handles "", null, decimals (rounded), and NaN, Inf and overflow. A field that can't be read now drops only the quota it defines, with a warning, instead of being stored as 0. Main also has a regression test for a Free account with a string pooledLimit.

Because main already defines UnmarshalJSON on CursorSpendLimitUsage and the fields are no longer plain int, this branch would not compile after a rebase, so I am closing it as superseded. If you still see a Cursor decode error on v2.14.14, please open an issue with the log line and a redacted payload and we will take a look. Thanks again!