Hey HN — solo builder here. I kept losing track of what competitors were shipping because checking 5-10 changelog/pricing pages by hand doesn't survive contact with an actual week. ShipDiff watches those pages daily, uses an LLM to classify what changed (feature/pricing/positioning) and rank how much it matters, and sends a Monday brief with a "why this matters for you" line per item, tuned to your product.

Technical notes since this is HN: it's a small pipeline (watch → classify → publish → onboard) running as three GitHub Actions cron jobs, SQLite as the only state (committed back to the repo each run — no database to host), and an LLM client with its own retry/backoff + provider fallback chain (Gemini → Groq → xAI → local Ollama) so a rate-limited API never blocks a run. Marginal cost per customer is close to $0.

$9/mo for 5 competitors, checkout collects your competitor list directly (no separate onboarding form). Would love feedback, especially on what "so what" actually means for your team — that line is the whole point and I'm sure I haven't nailed it for every use case yet.