Verastar
Verastar is a morning literature harness built by and for a busy vascular surgeon who can't read the hundreds of papers published in her field each month, yet can't afford to miss them.
Core trick
Claude Opus drafts literature summaries, then a deterministic verifier locates each statistic in source text and hides any number it cannot find.
Description
Verastar is a morning literature harness built by and for a busy vascular surgeon who can't read the hundreds of papers published in her field each month, yet can't afford to miss them.
The insight: frontier models are superb at reading papers, but can't be trusted on numbers. Verastar wraps Claude Opus in a deterministic verifier: every statistic is located in the source text and gated before it renders, showing what it proved, not what a model asserted. A number that can't be found is never shown; verified numbers appear inline. Papers link to their PDFs and full text, saved work forms a searchable library, and connections tie active projects to new literature.
Everything saved becomes a plain markdown file in a folder the clinician owns. Because that memory is an open folder, any agent she trusts can read it -- the seed of a research companion whole teams could build on. Verastar exists to bring doctors who haven't used AI into the fold, and give busy clinicians who never had time for research the chance to change their field.
Built end-to-end with Claude Code, including 181 adversarial false-verify tests, by a physician who never coded before late 2025.
README excerpt
# Verastar **A verifiable evidence digest for clinicians.** **▶ Demo:** https://youtu.be/RmT7JjTTqeM · **Live:** https://verastar.netlify.app *Vera* (truth) + *star* (your north stars): verified evidence, aimed at the concepts you steer by. You tell Verastar your *north-star concepts* and your active research projects. Every morning it hands you a digest of the last 1–3 days of literature that actually moves those north stars forward. Open the PubMed page, or the full-text PDF when the paper is open access. If a paper earns its salt, save it into your **Library**. The **Connections** view then surfaces the links between what you've read and the projects you're working on. And everything you save is written as plain markdown into a folder you own — your reading and your memory, on your own machine, never locked inside an app. It is the app version of a system I already run by hand: a personal knowledge base, a context portfolio, and a daily research briefing. Verastar makes it something any clinician can pick up — and I built it end-to-end with Claude Code, as a physician who never wrote a line of code before late 2025. Built for **Built with Claude: Life Sciences** (Anthropic × Gladstone Institutes), July 7–13, 2026 — **Build track**. --- ## The sacred core: verifiable, never fabricated Most "AI research digest" tools ask you to trust the model. Verastar does the opposite. Two non-negotiable rules govern everything: 1. **Quote-verification is the sacred core — built