Build automated repository health monitoring pipeline

Create a pipeline that checks the health of climate tech repos on a daily schedule and updates their status in the database.

For each project in climate_projects, the pipeline should:

1. Call GitHub API (/repos/{owner}/{repo}) to fetch: stargazers_count, pushed_at, open_issues_count
2. Call GitHub Issues API to count issues labeled good-first-issue
3. Check if a CONTRIBUTING.md file exists (/repos/{owner}/{repo}/contents/CONTRIBUTING.md)

Health score (0-100):

• Activity (0-40): score based on days since last commit. >90 days = 0, 0-7 days = 40, interpolate linearly
• Maintenance (0-30): issues with responses / total issues (cap at 30). Use comments count as proxy.
• Contributor friendliness (0-30): 15 pts if good-first-issue count > 0, 15 pts if CONTRIBUTING.md exists

Update climate_projects with { health_score, stars, has_good_first_issues, last_commit_at, health_checked_at }.

Infrastructure:

• Run as a Vercel Cron Job (/api/cron/health-check, daily at 02:00 UTC)
• Protect the route with CRON_SECRET header check
• Rate limit GitHub API calls: max 1 req/sec, exponential backoff on 429
• Authenticate with GITHUB_TOKEN env var (optional — degrades to 60 req/hr unauthenticated)

Write integration tests with mocked GitHub API responses covering: healthy repo, stale repo, rate limited response, missing repo (404).