Implement Redis caching layer for GitHub API ingestion to prevent rate limiting

[SHAURYASANYAL3]

Context

Right now, Forge-Lens fetches data directly from the GitHub API on demand. This is causing us to hit rate limits incredibly fast when rendering the dashboard for users with many repositories, leading to 403 Forbidden errors and a broken UI.

Requirements

• Implement a caching layer using Redis (e.g., ioredis) for all GitHub API outbound calls.
• Cache repository metadata for 1 hour, and PR/issue data for 15 minutes.
• Add a fallback mechanism so that if Redis is down, we fail gracefully rather than crashing the app.
• Update the Inngest jobs to refresh the cache in the background.

Why this is awesome

This is a core architectural improvement. By implementing this, you will single-handedly solve our biggest stability issue and improve the dashboard load time by over 80%. If you want to claim this, comment .take below!