Scott Belsky
LiveQA 64"Implications" newsletter, interviews, and The Messy Middle book
“The first mile of anything you build is where new users decide if you're worth their attention — most of you treat it as an afterthought, which is why most of you lose. The work itself happens in the messy middle, but the meta-skill is taste: knowing what to cut, what to keep, and when "interesting" beats "complete."”— Scott Belsky, in their own voice
One-command install isn’t live yet — the free download above is the full pack, working today.
8 AI skills included
Works with Claude, Cursor, Gemini CLI
Full knowledge graph access
Quality score
Independently judged, published as-is. This is how faithful the brain is to the real Scott Belsky.
Would they recognize themselves in it?
Their actual reasoning — no generic AI filler.
The failures and the mess, not just the highlight reel.
Claims sized to the evidence behind them.
How this score is computed
Every brain is graded by a panel of four independent AI judges — one per dimension — chaired by a calibration judge, against Scott Belsky’s own stated principles and public record. No judge sees another’s verdict. Hard numbers in the brain (revenue, dates, figures) are separately verified against the source corpus; any confirmed defect caps the score. We publish every score, including the low ones — they refresh with each QA run.
Explore this brain
Browse all 1086 atoms, connections, and thinking patterns. This is what you get.
8 thinking skills per brain
Interactive reasoning modes that chain into workflows.
/evolve
Watch a thinker change their mind across time — dated quotes, in order.
Learning/connect
Find a typed bridge between two ideas in the graph.
Creative/surprise
Curated atom you didn’t go looking for.
Creative/predict
Cascade second and third-order effects from their worldview.
Forecast/coach
Sequential Socratic questions in their voice.
Decision/debate
Stress-test a position from their actual stance.
Decision/teach
Concepts through their lens, in their frameworks.
Learning/advise
Counsel grounded in the brain’s atoms.
DecisionWhat's in the pack
Ask Scott Belsky something.
Same Claude model on both sides. The right side has Scott Belsky’s knowledge graph loaded.
Type a question above.
Same Claude model on both sides — one with a knowledge graph loaded, one without. The difference is the point.
Built from public sources only
This brain was derived exclusively from freely available, public resources — interviews, podcasts, free newsletters, public talks, and blog posts. No transcripts from commercial works or paywalled essays were used. Where book ideas appear, they come from public discourse: reviews, author interviews, press coverage, and widely discussed concepts.
Install in seconds
Claude Code / Cowork
Works today — download the pack right now.
Manual
Download the pack and drop brain-context.md into your AI tool's project context.