Institutions of One

Institutions of One is a research programme studying how people turn individual capability into durable, portable, ownable institutional power.

Live: https://ownership-platform.vercel.app

This repository is the programme's working research platform. It contains measurement instruments, validation operations, an evidence-governed Observatory, longitudinal research infrastructure, public research surfaces, and participation and partnership workflows.

The project is broader than a creator quiz. The Ownership Index is one instrument within the programme, and the Portfolio Professional instrument is a related but distinct line of inquiry. Neither candidate instrument should be described as validated, normed, or activated until its recorded research gates are met.

Public functions

The platform is being organized around five public jobs:

1. Understand the phenomenon — explain the Institutions of One theory, constructs, limits, and research questions.
2. Measure yourself — provide clearly labeled candidate instruments and useful results without overstating validation.
3. Explore the evidence — publish reviewed Observatory cases, sources, relationships, and longitudinal events.
4. Follow the research — release methods, findings, corrections, briefs, and articles.
5. Participate or partner — support research participation, nominations, institutional partnerships, and responsible commercial inquiries without compromising research independence.

Core research constructs

The programme must keep these related concepts analytically distinct:

• Capability — what a person can reliably do.
• Portability — whether that capability and its evidence travel across employers, platforms, clients, and jurisdictions.
• Ownership — which audiences, rights, assets, data, relationships, and revenue channels the person or their entity controls.
• Institutionalization — the extent to which capability has been converted into repeatable systems, roles, records, governance, and continuity beyond a single act of labor.
• Authority — recognized permission or legitimacy to decide, convene, publish, transact, or set standards.
• Durability — whether the system survives shocks, absence, platform change, or personnel turnover.
• Optionality — the number and quality of viable paths available without dependence on a single gatekeeper.
• Distribution — the ability to reach people and move work, ideas, products, or opportunities.
• Capture — who retains the economic, reputational, informational, and strategic value created.
• Leverage and economic power — the capacity to produce effects disproportionate to direct labor and to negotiate from control rather than dependence.

These definitions are the current theoretical baseline. Instrument mappings and empirical relationships remain research questions until supported by the validation programme.

What is implemented

• A pure, typed Ownership Index scoring engine and versioned item bank.
• Interactive assessment and stateless assessment API.
• Evidence tiers, immutable score-event architecture, and versioned methodology records.
• Candidate-instrument validation operations, including participant records, cognitive interview sessions, response-process coding, revision records, gate decisions, pilot protocols, and activation decisions.
• An Observatory evidence model separating cases, claims, sources, relationships, events, construct observations, audit history, and reviewer packages.
• Consent, privacy, nomination, research-subscription, and partner-inquiry infrastructure.
• Longitudinal and publication-oriented data foundations.
• Automated tests and production migrations in the deployment build.

Current status

The infrastructure is substantial, but the research programme is not yet complete.

• Published scoring remains provisional and must not be represented as validated or normed.
• Candidate-instrument validation and activation gates remain controlling.
• Observatory dossiers remain draft unless their review and publication states say otherwise.
• Cross-case findings require a deliberately sampled, sufficiently reviewed cohort.
• The public theory, methodology, findings, and participation journey still require final integration and end-to-end launch verification.
• The MrBeast evidence thread is frozen. No additional subject-level research should displace whole-program completion work.

Current completion sequence

Future work should follow this dependency order:

1. Lock and publish the theoretical architecture and construct boundaries.
2. Reconcile the Ownership Index and Portfolio Professional instrument purposes, mappings, overlap, evidence requirements, version states, and validation gates.
3. Align the public information architecture to the five public functions above.
4. Establish and complete a bounded, representative first Observatory cohort.
5. Produce the first defensible cross-case findings and research publications.
6. Verify the complete production journey, including consent, privacy, accessibility, mobile behavior, database operations, corrections, analytics, and deployment state.

This sequence—not the nearest unfinished case dossier—governs future “proceed” instructions.

Local development

npm install
npm run dev
npm test
npm run build

The production build runs database migrations, tests, and the Next.js build.

Deployment

Production deploys only through commits pushed to main and the repository's Vercel integration. Do not deploy with the Vercel CLI.

Release gate

data/releases.json is the one record of what is public. Each edition, case record and tool (including the interactive tools) has one entry: id (the same id across RN's repos, such as ioo/edit-005 or ioo/case-mrbeast), releasedAt (a date in Pacific/Honolulu, or empty), hold, postUrls, longFormUrl, sourcePageUrl, and dateSource (where the date came from). A piece is public only when releasedAt is today or earlier and hold is false. Anything else returns 404: its page, its place in every list, the sitemap, its share image and its files, including through the image optimizer. The related list holds the Public Reader visual stories this site links to, with the dates from rn-selected-work data/sources-ioo.json; a link to one appears only from its date.

• The rule lives in src/lib/release-rule.mjs. next.config.mjs fixes "today" to the build day in Hawaii and inlines the list of public ids, so pages, browser code and src/middleware.ts agree and nothing about a held piece reaches the browser. The /tools routes (page.tools.tsx) join the build only once a tool is public.
• npm run build ends with scripts/check-releases-built.mjs, which fails the build if a held piece has a page, a sitemap entry, or its title or address in any served page or script, or if a released piece is missing its page, sitemap entry or links (posts, source page, long-form, gallery index, previous and next).
• .github/workflows/daily-release.yml runs at 00:05 in Hawaii. On a release day it pushes an empty commit to main, and Vercel rebuilds as for any push. npm run releases:today shows what goes live today.
• Review renders (RELEASE_REVIEW=1 or CYCLE01_RENDER_FREEZE=1) show held pieces for checking by eye. They never apply to the production deployment.
• To release a piece, set its releasedAt and hold: false. Never guess a date: a piece without a verifiable date stays held with an empty releasedAt.
• Pieces include the two interactive tools already on the site (kind: "interactive"): the Dependency Explorer (/observatory/dependencies) and the Ownership Index (/assess/creator). Each carries the link set and previous and next between the two.
• Every released piece has a sources page on RN Collins, selected work, at https://rn-selected-work.vercel.app/sources/ioo/<id after ioo/>, and links it (sourcePageUrl). The pages come from this repo: scripts/export-selected-work-sources.ts writes rn-selected-work data/sources-ioo-site.json from the edition source lists, the case records' sources and the release dates here (how to run it is in the script). When you release a piece, run it, commit that file in rn-selected-work, run its npm run releases:sync, set the new entry's galleryUrl and postUrls there, and set sourcePageUrl here. The build check fails while a released piece has no sources page address.
• Long-form: an edition is its own long-form piece; a case discussed in an edition points at that edition; any other case record is its own long-form piece and says so on the page.

Optional services

• Postgres / Prisma — research records, evidence, consent, reviews, and immutable events.
• Upstash Redis — public API rate limiting.
• Resend — transactional and research communications.

See .env.example, MIGRATIONS.md, and the source-level architecture rules for configuration and operational constraints.