What should you pick back up when a year of projects, notes and conversations is scattered across folders? This prototype starts with a smaller question it can actually answer: what files are there, and which are exact copies?
The working slice is a read-only census and a local dashboard. The public demonstration counts four bundled synthetic files, identifies one duplicate group and shows those facts on screen. Choosing today's project is still an idea, not a feature.
Node.js 22.13 or newer and Python 3.11 or newer.
npm ci
npm run demo:census
npm run build
npm startOpen the loopback address printed by the preview server. For live editing, use npm run dev -- --hostname 127.0.0.1.
The census reads two fixture folders containing four files. Two Markdown notes are exact copies, giving one duplicate group and zero read errors. The Sources panel displays those computed counts. The command accepts no arbitrary source path: public build data can only come from the bundled synthetic fixture.
Follow one complete census, from files to screen: the four inputs, hashing, duplicate grouping, local outputs and counts-only projection. The mechanism notes link each step to its implementation.
This prototype does not implement semantic indexing, ranking, recommendations, external research or the proposed Work/Threads/Timeline views. Metadata counts are not a basis for pretending those insights exist. Real-source configuration is local-only; do not copy personal inventory JSON into the public app. See docs/SAFETY_CONTRACT.md for the source boundary.
npm test runs the read-only inventory tests and builds/server-renders the dashboard.
Configuration rejects non-object JSON and requires a real JSON boolean for include_hidden; the string "false" cannot silently enable hidden-file scanning.
The original idea was a daily docket with attributed conversations, project history and evidence behind each suggestion. Future directions show the missing steps and how each could be demonstrated before calling it a feature.
MIT licensed; see LICENSE.md. Origin and release boundaries are documented in ORIGIN.md and SECURITY.md.
Open the saved synthetic result alongside its input and demonstration. The result is from the bundled synthetic example; local machine paths and temporary run identifiers are excluded from public projections.