Domain Intelligence / Investing Operations · Archived
Knowing when not to buy.
Completed research into evidence-driven domain discovery and guarded purchasing. The experiment found no candidates that justified buying under its policy, and scheduled scans were stopped.
Saved product captures. Prices, example records and interface state are shown as captured; they are not current offers or a live status feed.
Domain Scout collected live technical signals, generated candidate names, checked availability and prior history, and evaluated modeled expected value behind deterministic spend gates. The completed experiment is documented as a negative result: available candidates did not justify buying under the standing assumptions. The repository is not maintained as a product for outside use, and scheduled scans were switched off on August 22. Historical screenshots show the research interface, not an active investment service. The result highlights uncertain lookup states, unvalidated sale-probability assumptions and the need to stop automation when its premise fails.
A completed research experiment in domain discovery, modeled value and spending controls.
Pending candidates with theme, rationale, quality score, trademark-risk band, availability context, and source tags. Decision brief shows confidence and purchase panel controls without enabling autobuy.
Watch status filtered into its own tab so operators can park names without approving spend intent prematurely.
On-demand registrar checks with database-backed rate limiting and clear unavailable / retry states.
Connector cards for live, needs_auth, needs_key, permission_required, and disabled states with last-checked times.
Detect → Generate → Challenge → Learn stages plus checklist gates for availability, price, sources, trademark risk, natural speech, and confidence. Autobuy eligibility shown as currently disabled.
Aggregates scan runs, candidates, availability checks, and paper spend with calibration requirements. No simulated rows are inserted by the app.
Idempotent porkbun-manual draft after confirmed live purchase; operators paste marketplace URL and record inquiries, offers, sale price, fees, and drops. Marketplace listing creation remains manual.
Readiness checklist for database, live sources, registrar credentials, LLM gateway, live purchase capability, and autobuy-disabled confirmation, plus server-enforced money limits.
Token login issuing an HttpOnly session cookie; APIs also accept bearer admin token. Not a public self-serve product.
| Platform | Web |
|---|---|
| Availability | Archived |
| Source record | October 2026Synchronization date, not a capture or launch date |
| Who it is for | Builders evaluating automated research systems · Teams studying evidence quality and spend controls |
| Tools behind the build | Next.js · React · TypeScript · Tailwind CSS · Neon Postgres · Vercel · Vercel AI Gateway · Porkbun · Zod |
Recorded update · 2026-08-22
Domain Scout is now documented as a completed negative result rather than a product for outside use. Its expected-value policy did not justify buying the available candidates, and scheduled scans and quote sweeps were switched off.
| August 2026 | Experiment concluded; scheduled scans stoppedDomain Scout is now documented as a completed negative result rather than a product for outside use. Its expected-value policy did not justify buying the available candidates, and scheduled scans and quote sweeps were switched off. |
|---|---|
| August 2026 | Domain Scout published, with the spend gates documentedThe scout board, live source connectors, prediction engine, paper-trade lab, acquisitions pipeline and launch control are now documented publicly. Live purchasing stays off by default behind paper mode, a kill switch, budget and price ceilings, and exact confirmation. |
Start with the business task. We will work out what needs building and what belongs in the scope.
Thursdays
Each Thursday: the shop notes and AI daily issues from the last seven days, and one product. Unsubscribe from any issue.
