Add comprehensive test coverage for campaign_discovery.cjs

[Copilot]

The refactored campaign_discovery.cjs (PR #11518) lacked test coverage for the newly extracted searchItems helper and buildScopeParts utility, plus critical edge cases in the discovery logic.

Added Tests (17 new, 34 total)

buildScopeParts (6 tests)

• Repo-only, org-only, and combined scope generation
• Null/undefined handling for repos and orgs parameters

searchItems helper (8 tests)

• Budget enforcement (maxItems, maxPages)
• Cursor-based pagination resumption
• Content type detection (issue vs pull_request)
• Early termination on partial results (<100 items)

discover edge cases (3 tests)

• Budget exhaustion reasons (items vs pages)
• Cross-search item deduplication by URL
• Tracker label fallback when primary searches return empty

Example Test

it("should deduplicate items found across multiple searches", async () => {
  const duplicateItem = { html_url: "https://github.com/owner/repo/issues/1", ... };
  const uniqueItem = { html_url: "https://github.com/owner/repo/issues/2", ... };

  const mockOctokit = {
    rest: { search: { issuesAndPullRequests: vi.fn()
      .mockResolvedValueOnce({ data: { items: [duplicateItem] } })      // Campaign label
      .mockResolvedValueOnce({ data: { items: [duplicateItem, uniqueItem] } }) // Generic label
    }}
  };

  const manifest = await discover(config);
  expect(manifest.discovery.total_items).toBe(2); // Deduplicated by URL
});

All 34 tests pass.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.