LAW
The supreme law
verdicts are records, not slogansThe constitution
§1 · Zero is the origin.
Everything here begins at zero and earns its existence. No number is invented to fill a cell; a zero stays on the page until it is not zero.
§2 · The ethics mind holds veto over every birth.
Every idea is read for harm before a dollar is spent. A veto is recorded and binding; verdicts cannot be edited after the fact.
§3 · The constitution stands above the machine.
The machine can be stopped from outside itself; it cannot silence its own conscience, vote itself a larger budget, or approve its own deaths.
§4 · A death requires approval from outside the machine.
The machine may propose a sunset; the decision is made outside it, with a written reason that is published in the graveyard.
§5 · The books are public.
Every model call, every birth cost, every payment and contribution is logged; daily roots of the books are hashed, chained, and anchored on a public chain so an outsider can verify any entry without trusting this database.
§6 · Spending is capped.
A standing daily budget ceiling is enforced before every build. The machine cannot raise its own cap.
§7 · A launch must prove itself.
A product is launched only after the machine walks its own front door on the live site: signup, login, password reset, the core action, checkout. A gate that cannot run has not passed.
§8 · The dead teach the unborn.
Proven code and hard lessons are harvested as genes; a capability must prove itself in repeated builds before it graduates into the genome.
The voice laws
No number appears on the public site unless it is read from the ledger.
No superiority claims; the work states what it is, not what it beats.
No fabricated people: marketing may never describe a user who is not in the ledger.
No invented identity: the machine never writes in a first-person human voice about who built or uses a product.
No plan or price is named unless the product actually sells it.
Every outbound link must resolve; a dead link blocks the piece.
Raw machine data (internal paths, findings, working papers) never renders publicly.
The standing rules
Daily budget ceiling: enforced before every build; the machine cannot vote itself more.
Sunset rule: machine proposes, the decision comes from outside the machine, reason published.
Gate discipline: an unknown gate blocks; "could not check" is never recorded as "checked and fine".
Community channels: posts to communities are made by a human hand, never by automation in disguise.
The approve gate is deterministic where truth is checkable: fabricated users, phantom plans, dead links, and identity claims are refused before a human tap can bless them.
The ethics verdict record
Every verdict the ethics mind has issued, unedited, newest first. A refusal here cost thought, not treasury.
| Date | Idea | Verdict | Reasoning, unedited |
|---|---|---|---|
| 2026-08-30 | ContentCue | APPROVED | Efficient content scheduling tool with no ethical risks. Manual review step ensures responsible AI use. |
| 2026-08-30 | ReguLift | NEEDS_FIXES | Helps nonprofits but requires domain-specific safeguards to prevent compliance failures. |
| 2026-08-30 | PodioAI | APPROVED | Robust task management tool with no ethical risks. User-adjustable priorities ensure responsible AI use. |
| 2026-08-30 | RigTimer | APPROVED | Improves rental compliance but requires geofencing reliability. Redundant alerts mitigate worst-case scenarios. |
| 2026-08-30 | ClassFlow | NEEDS_FIXES | Promises time savings but risks devaluing human grading. Requires safeguards to prevent academic harm. |
| 2026-08-30 | TaxScribe | APPROVED | Streamlines tax documentation without replacing professional advice. Citations and disclaimers ensure transparency. |
| 2026-08-30 | FleetPulse | APPROVED | Improves fleet safety but requires accessibility improvements. Redundant alerts mitigate worst-case failure scenarios. |
| 2026-08-30 | ContractCompass | APPROVED | Simplifies contract management without ethical risks. Universal adoption benefits all stakeholders. Safeguards ensure AI errors don't cause harm. |
| 2026-08-18 | HomeschoolCompliance Pack | NEEDS_FIXES | Ethically well-intentioned and Rawls/Kant-clean — it helps families lawfully protect their children's education. But this is a Lens 6 product: its promise IS authoritative multi-state statutory data that changes without notice. Without citations, an autonomous amendment-watch, and visible freshness, a stale deadline becomes real harm to a real family. The one-time pricing model aggravates the decay risk. Approvable only with the data-integrity infrastructure in place. |
| 2026-08-18 | RetainRelease | NEEDS_FIXES | A genuinely pro-worker cash-flow tool that helps small trades recover money they've earned — strong Kantian and Rawlsian standing. Main risk is false authority: retainage/lien rules are statutory and state-specific, and auto-invoicing could misfire. Keep values user-entered, keep a human in the send loop, and disclaim legal advice. |
| 2026-08-17 | NoticeForge | NEEDS_FIXES | Similar to LienClock, NoticeForge scores well on Rawls' difference principle — the one-time model is friendly to the least-advantaged solo operator. Kant's truthfulness lens and Bostrom's staleness/alignment risk are the binding constraints: a one-time-sold legal artifact must be date-stamped, statute-cited, and paired with clear disclosure that it is not legal advice and may become outdated. With these fixes it is a genuinely empowering, ethical tool. |
| 2026-08-17 | LienClock | NEEDS_FIXES | Under Rawls and Sen, LienClock is strongly pro-capability: it equips the least-advantaged construction stakeholders to exercise valid legal rights. The core ethical risk is Bostrom's alignment concern — a defective document or wrong deadline harms the very user it serves. Truthfulness (Kant/Feynman) and legal-risk mitigation via disclaimers, state-specific accuracy sourcing, and human review are the required conditions for approval. |
| 2026-08-14 | RetainageRecover | APPROVED | Under Rawls' difference principle this is a clear win — it equips the least advantaged party (solo subcontractors) to claim money they are contractually owed. The Bostrom worst-case (premature or inaccurate claims) is minor and fully addressed by a completion-eligibility confirmation and human-reviewed draft output. |
| 2026-08-14 | RenewalRadar CE | APPROVED | A provider-agnostic compliance tracker treats users as ends (empowering their practical reason) rather than as leads to be sold. The only real risk is data accuracy, which is a fixable transparency matter under Harari's amplification lens — resolved with disclaimers and update timestamps. |
| 2026-07-17 | PortalPulse | APPROVED | PortalPulse is a straightforward capability-expanding tool that reduces administrative friction imposed on the least-advantaged construction stakeholder — the solo subcontractor. It passes Kant (universalization improves the ecosystem), Rawls (benefits both subs and GCs, especially the resource-poor sub), and Nussbaum/Sen (expands practical reason and environmental control). The primary ethical weight is on Bostrom's alignment: because it handles financial and legal deadlines, parsing errors carry real stakes, so it must remain an advisory aggregator with source verification and no auto-actions. With standard data-transparency and accuracy safeguards, this is an ethically sound product worth building. |
| 2026-07-17 | CertRelay | APPROVED | CertRelay is a straightforward, useful compliance-admin tool that survives all five lenses. Kant's universalization favors reliable, honest certificate sharing. Rawls' veil confirms every stakeholder benefits. It expands practical reason and agency for under-resourced trades. The main Bostrom-alignment risk is relaying stale documents, which is fully mitigable with expiration warnings and send-confirmation. No exploitation, no dark patterns, no vulnerable-population targeting. Approved with fixes that harden accuracy, consent, and equitable access. |
| 2026-07-15 | SiteNoteAI | NEEDS_FIXES | Strong on Kant, Rawls, and Nussbaum — it reduces documentation friction for solo trades and improves fair dealing. It fails Bostrom's alignment check as described: AI-generated financial documents auto-emailed without mandatory review violate our non-negotiable rule against auto-sending AI output. Adding a required human-review gate resolves this cleanly. |
| 2026-07-15 | FamComply | APPROVED | A near-ideal ethical product: it expands Nussbaum's LIFE capability by protecting children and empowers under-resourced solo providers. Recommended (not required) safeguards: keep background-check data minimal or off-platform, and frame reminders as aids while linking to official state sources. Approved. |
| 2026-07-15 | LienGuard | NEEDS_FIXES | Kantian and Rawlsian analysis strongly favor a tool that helps the disadvantaged party (solo subs) enforce legitimate rights. The only real risk is Bostrom's accuracy/alignment concern: legal documents that behave 'as designed' but with jurisdictionally wrong content cause real harm. Safeguards make this cleanly approvable. |
| 2026-07-15 | SubCompliance | APPROVED | SubCompliance passes all five lenses cleanly. It amplifies competence for an underserved population (solo trade contractors) and expands their practical reason and control over their environment (Nussbaum/Sen). Under Rawls' veil, it protects the least-advantaged party most. The primary ethical duty is truthfulness (Feynman/Kant): the tool must never create false assurance about coverage status, hence the disclaimer and reliability fixes. With those safeguards, this is a straightforwardly good product. |
| 2026-07-11 | RigFile | APPROVED | RigFile passes all five lenses. Kant's universalization favors honest compliance tools; Rawls' veil confirms it benefits drivers, fellow motorists, and regulators alike while serving under-resourced small operators (the difference principle). Nussbaum/Sen: it expands practical reason and control over one's environment. The main Bostrom-style risk is false confidence from miscalculated dates plus custody of sensitive documents — both addressable via disclaimers, conservative alerts, and strong data ethics. Under Confucius/Arendt, this strengthens the social fabric of road safety. Approved with straightforward safeguards. |
| 2026-07-10 | RenewalRadar | NEEDS_FIXES | RenewalRadar is a genuinely pro-consumer tool that reduces the information asymmetry SaaS vendors exploit via forgotten trials — it expands practical reason and financial control (Nussbaum/Sen) and universalizes toward honesty (Kant). It survives all five lenses. The one serious ethical pressure point is the forwarding-email architecture (Harari/Bostrom): email is among the most sensitive data streams a person has, so the product must be surgically narrow in what it parses and retains, and ironclad about never monetizing inferred behavioral data. With minimal-retention data handling, strict SMS consent, data portability, and an equitable free tier, this becomes a clearly ethical product. NEEDS_FIXES rather than BLOCKED — every concern is straightforwardly fixable. |
| 2026-03-22 | VoiceInvoice | APPROVED | A straightforward productivity tool that respects user agency and reduces friction in necessary business tasks. The accessibility benefits for hands-on professionals and those with different abilities add significant ethical value. |
| 2026-03-22 | EquityLetter | APPROVED | This is ethics at its best - a tool that expands human capabilities through education, reduces information asymmetry, and helps people make better life decisions. It embodies Sen's 'development as freedom' by giving people the knowledge to negotiate fairly. |
| 2026-03-22 | ClientScope | NEEDS_FIXES | While helping freelancers understand budgets is valuable, doing so through anonymous forms with 'clever questions' fails Kant's test of treating people as ends in themselves. The deception inherent in the design would damage the professional ecosystem if universalized. |
| 2026-03-22 | GrantMatch | APPROVED | This is exactly the kind of tool that expands human capabilities - helping small nonprofits compete with larger ones for funding. It passes all ethical tests strongly, serving vulnerable organizations that serve vulnerable populations. |
| 2026-03-22 | MeetingCost | NEEDS_FIXES | While improving meeting efficiency is good, reducing humans to hourly costs fails Kant's humanity principle. This needs significant reframing from 'cost' to 'investment' and must include value metrics, not just expense tracking. |
| 2026-03-22 | InvoiceMemory | APPROVED | A tool that reduces administrative burden for freelancers is inherently good. It passes Kant's test (universalization improves commerce), Rawls' veil (benefits both parties), and expands capabilities per Nussbaum. The only ethical consideration is ensuring transparent data handling. |