An organization with a documented failure history and certified infrastructure is more trustworthy than one with a clean record and no ledger. Certification verifies you meet the FROZEN-2.0 standard — and maintains that verification through a cryptographic record that cannot be quietly amended.
For organizations: Every AI failure deserves a permanent, immutable record. Log it. Seal it. Append the remediation. The record cannot be edited after the fact. The AION-Registry holds public certification outcomes.
For individuals: Write the exact present-moment observation. Seal it with a triple-time cryptographic stamp. The SHA-256 seal is permanent. Tamper-evident. Requires no audience.
Same mechanism. Same FROZEN-2.0 stamp function. Same SHA-256 seal. Different stakes.
The triple stamp is not redundancy. It is a claim: this moment of human significance deserves to be held simultaneously in every major civilizational framework for measuring when.
Tiers 1 and 2 submit anytime — payment via Stripe confirms intake automatically. Tiers 3 and 4 are architect-led. Intake Monday and Tuesday only. Submissions on other days are voided and non-refundable. Delivery: weekends for all tiers.
Every certified organization receives an Epistemic Debt Score (EDS) — a 0–100 metric measuring AI epistemic integrity across five components. Formula is public.
10-audit-request.yml
10-audit-request.yml
13-integrity-violation.ymlEvery certified organization receives an Epistemic Debt Score — a 0–100 metric measuring AI epistemic integrity across five components. The formula is public. You can read the spec, audit the computation, and verify the score independently. A score is not a pass/fail. It is a measure of how much unacknowledged epistemic liability your AI deployment carries.
Structured submissions are made through GitHub Issues using the official template set. Each template produces a structured, legally-declared record before sealing. Blank issues are disabled. Every submission uses a template — every template carries native file upload, SHA-256 binding, declaration checkbox, and legal compliance language.
| ID | Template | Use Case | Note |
|---|---|---|---|
| 01 | ai-failure | AI system output failure — identity verified, legal declaration required. Pre-remediation seal. | Declaration |
| 02 | research-priority | Hypothesis or finding sealed before results are known. Kills HARKing. No institution required. | SHA-256 |
| 03 | evidence-chain | Document or source communication at point of receipt. Chain-of-custody proof. | SHA-256 |
| 04 | creative-priority | Creative work authorship — music, writing, art, design, code. Cryptographic priority proof. | SHA-256 |
| 05 | clinical-record | Clinical incident, trial data, or institutional record. PHI gate active. FDA 21 CFR Part 11 compatible. | PHI Gate |
| 06 | scope-anchor | Agreed project scope or negotiated position before work begins. Scope creep disputes resolved by the ledger. | Declaration |
| 07 | general-trace | Any observation, decision, or record that doesn't fit another template. Universal entry point. | SHA-256 |
| 08 | foresight-seal | Dated professional prediction before it resolves. Cryptographic foresight track record — cannot be fabricated after the fact. | SHA-256 |
| 09 | webeater-link | Cryptographic link between two sealed entities — binds a new seal to an existing SHA-256. | SHA-256 |
| 10 | audit-request | Request a certified audit. Stripe payment required at submission. Automated intake for Tiers 1–2. | Stripe |
| 11 | audit-completion | STP Certified Auditor files a completed audit to the ledger. Auditor badge verified live on submission. | Auditor Only |
| 12 | auditor-application | Apply to become an STP Certified Auditor — skills-based vetting, no credentials required. | Application |
| 13 | integrity-violation | Report badge misuse, bribery, or coercion. Permanent ledger record. Cannot be retracted. | Permanent |
| 14 | near-miss | AI output that almost caused harm — caught before impact. Near-miss reporting before internal review begins. | Declaration |
The SHA-256 proof is the same whether the content is an AI failure report, a research hypothesis, a hospital incident record, a contractor's agreed scope, or a professional foresight declaration. The stake determines how you use the sealed record afterward.
The mechanism and the auditor ecosystem are two sides of the same integrity guarantee. The stamp function does not change. The auditor network extends reach without diluting standards.
sovereign_trace_stamp.py is FROZEN-2.0. Written once, verified once, deployed permanently. No patches. No updates. The stamp it generates is only permanent if the code that generates it is also permanent.stamp/FROZEN-1.0-RETIRED/. That disclosure is in the ledger.
.github/verified-auditors.json on every submission..github/revoked-auditors.json. The revocation record is permanent. A revoked badge cannot be reinstated."Responsible AI" without failure documentation is marketing. The Sovereign Trace Protocol exists because the honest record — logged before remediation, sealed before dispute — is the only record that matters when accountability is tested.