For every essential medicine, we trace the full chain — finished drug to active ingredient to the key starting materials (KSMs) it's built from — show where that chain is dangerously concentrated, and find credible routes to make it in the USA.
Nobody can fix a supply chain they cannot see.
Phase 1 today: the medicine backbone — WHO essential medicines, ATC codes and the EO 13944 US-priority overlay — all real and browsable now. Supply-chain mapping, synthesis routes and the alternative-drug finder shown here are illustrative previews until later phases populate them.
The data is a graph, not a table — because the most important answer is often a different drug entirely. We trace each medicine across all three.
We break each active ingredient down into its key starting materials, recursively, and map who makes them and in what country. Single-supplier and single-country links surface as risk.
A retrosynthesis question, computed with open tools and constrained to US-sourceable building blocks. Proposed routes are always labeled as computational — never mistaken for established commercial reality.
A graph traversal: from a blocked drug, across the conditions it treats and how it works, to other FDA-approved drugs whose supply routes aren't blocked.
The same five states label every node, list, and badge across the atlas. Color is always paired with a distinct shape and a word, so nothing depends on color alone.
This is a sincere, non-partisan resilience project, built in the open. Whatever your field, there's a good first issue with your name on it.
Static-site tooling, graph rendering, the DuckDB-WASM query layer, data exporters. The whole stack is open.
Good first issuesReview extracted synthesis routes, validate key starting materials, and sanity-check computationally proposed alternatives.
Review queueTrace manufacturers, surface single-country concentration, and identify domestic candidates that could make a target.
Manufacturer tasks