Build: Cloud Reporter — Milestone 2

You measured Milestone 1's sequential fetch. Now make it concurrent — with the exact tools from this module — and prove the output didn't change.

The assignment

1. Discover, then fan out. First request learns how many pages exist (Link header, or fetch until short page once and remember). Then fetch all remaining pages concurrently: one goroutine per page, results into a channel, sync.WaitGroup to know when you're done.
2. Identical output. Collect everything, then sort — result order must not depend on goroutine timing. The report must be byte-identical to Milestone 1's (pipe both to files and diff them).
3. Race-clean. go test -race ./... — every mistake this module's "Common Mistakes" section warned you about will attempt to happen in this code.

Done when

• diff <(git stash && go run . repos --org kubernetes --format json) <(git stash pop && go run . repos --org kubernetes --format json) — or just save the old output first — shows no difference.
• go test -race ./... is clean.
• The wall-clock number you wrote down in Milestone 1 just dropped by roughly the page count. Update the comment with the new time — that ratio is this module's lesson in one number.

Commit.