status_authority: NONE
Purpose: Outward-facing and durable human-readable materials that are not the control plane, not experiment evidence, and not product marketing as trust.
| Kind | Examples | Path habit |
|---|---|---|
| Statements | Readiness declarations, public posture, press notes | statements/ |
| Publications | Mandatory end-of-round seat reports (what / why / verified); cumulative evolution log | publications/ |
| Epistemic analyses | Mandatory seat-on-seat + operator performance as INFERENCE; self-gap | publications/epistemic/ |
| Website design | White paper + tech spec for distrust-first public surface | website/ |
| Tutorials | How to enter the lab, how to run doctor/verify, how to file a return | tutorials/ |
| Docs for humans | Plain-language explainers that may later leave the repo | docs/ |
| Press / external | Drafts for journalists, partners, public posts | press/ |
| Templates | Empty shapes for future comms | templates/ |
At the end of every real work round, every AI files a report per
Agent Resources/Tools/Skills/essential/round-close-publication.md
into publications/rounds/ and updates publications/INDEX.md.
This is not STATUS authority and cannot soft-close reds.
- Secrets, credentials, unredacted private identity
- Live keys, wallet seeds, or anything that implies real-world economic value
- Soft-closing of security reds (T-01, CARD-11, etc.) by eloquence
- Claims that multi-AI agreement is independent science
- Replacement of
STATUS.json/NEXT_ACTION.mdas operator authority
- Constitution +
STATUS.json/NEXT_ACTION.md - Load-bearing distrust:
WHY_NOT_TO_TRUST_THIS_PROJECT.md - Evidence receipts under
operations/receipts/ - This folder — communication, not promotion of status
A polished statement here cannot outrank a red on main.
statements/2026-07-22_READY_FOR_SERIOUS_RESEARCH.md
Creating a communications tree is not a product launch, not a token, not a security certificate, and not a waiver of permanent distrust reasons.