How the articles on this site are researched
The data-heavy articles on this blog draw on hundreds of deep research sweeps across 15+ sources. Here is the stack, the sources, and what has been surfaced so far.
The engine
The research engine is Jester — a 20-source intelligence engine that scans Reddit, HackerNews, IndieHackers, GitHub, ProductHunt, Substack, arXiv, YouTube transcripts and more in parallel, then synthesizes the findings into a structured brief: clusters, contradictions, money signals, niche gaps, source citations.
A free hosted version runs in the browser at /jester. Anyone can drop in a topic and see the same kind of brief that anchors the articles here.
Sources scanned
- Reddit (subreddit search + comment scoring)
- HackerNews (front page, Show HN, comment threads)
- GitHub (repo stars, issues, recent commits, README parsing)
- IndieHackers (post threads + revenue posts)
- ProductHunt (launches + comment sentiment)
- Substack (newsletter signal)
- YouTube (transcripts via Whisper)
- X (post threads when accessible)
- arXiv (research frontier papers)
- Lobste.rs (developer discourse)
- Amazon (bestseller rankings, review patterns)
- CoinGecko + Solana on-chain (when crypto-relevant)
- Exa (neural web search)
- Firecrawl (page fetching at scale)
- Vendor blogs and changelogs (direct primary sources)
Coverage so far
Research reports archived to date, by cluster:
- MCP servers & Claude Code52 reports
- Crypto trading & Solana memecoin20 reports
- Affiliate sites & monetization12 reports
- Self-publishing & memoir niches10 reports
- Faceless YouTube & video automation8 reports
- AI research tools & GEO patterns8 reports
- Indie hacker pain points10 reports
- Vibe coding workflows6 reports
- Freelance & Upwork dynamics4 reports
How an article gets written
- Topic chosen, scope defined.
- Jester runs a deep sweep across the relevant 5-15 sources.
- Findings cluster automatically: high-conviction, multi-source, single-source, contradictions, gaps.
- Numbers, quotes, named tools and primary links pulled into a working brief.
- Draft written from the brief, every claim traceable to a source.
- Voice and structure edited by hand.
- Published.
Why this is public
Most blogs do not show their work. The cost of that is that readers cannot tell which claims are real and which are vibes. The articles here cite sources and lean on volume of inputs over personal opinion. The engine that makes it possible to do this at the cadence of a daily-building solo project is the same one you can use in your browser at /jester.