Launch Watch / Project note
A focused queue for launch signals
Launch Watch turns official Hacker News lists into a bounded private review queue, retaining recent eligible sampled stories while reserving selective alerts for items with both tool related topics and request or feedback signals.
- Our role
- Bounded research collection and review queue
- Stage
- Internal tool
- Category
- Tools & automation
Launch Watch workflow
Sample Official Lists
Read official Show HN and Ask HN API lists
Retain Eligible Stories
Keep recent valid stories and deduplicate IDs across runs
Bound Private Queue
Apply retention and item limits while preserving explicit errors
Alert Selective Signals
Show limited links when topic and request signals align
Workflow illustration from source, not live UI, showing official lists, eligible sampling, ID deduplication, bounded queue storage, and selective human review.
Make Signals Reviewable
Launch signals are useful only when people can review them in context. We designed a focused collection workflow that reduces repeated source checking, preserves recent eligible stories, and keeps selective alerting separate from queue eligibility and any assumption of demand.
Separate Queue From Alerts
We sample official Show HN and Ask HN API lists, keep recent valid timestamps, and deduplicate item IDs across lists and runs. Retention and item limits bound the private queue, while normal output stays silent unless both topic and request signals are present, with up to three links for context.
Review Stays Focused
The workflow gives a reviewer a current, bounded set of sampled stories without implying that every source item qualifies. Selective alerts add context when both signals align, while each record remains outside outreach and fixed to email_eligible=false.
Bounded review design
The source-backed design separates recent eligible queue retention from selective alerts, using official lists, deduplication, bounded storage, explicit errors, and limited context links.