EncrypiaBrief companion v0.2September 2026Public · Pre-testnet
01 · Primer · Whitepaper
The problem, and what an agreement looks like.
Treat this reading as a primer: you should leave with a walked object — a typed agreement — not only a slogan. Encrypia publishes this companion to public technical brief v0.2. Zeqron is the protocol. September 2026. Status: advanced design and substantial software in a Rust monorepo. No measured public network. Architecture · Pre-testnet.
Most distributed books still solve a 2008 problem: one global list of transfers. Every write competes for the same state machine. The cryptography that signs that list — ECDSA, Ed25519 — is already on NIST’s replacement schedule under FIPS 203, 204 and 205. Migrating a live network later is a multi-year systemic risk: what is stored today can be read tomorrow.
That model leaves three gaps. First, global-state contention: scaling becomes a permanent tax. Second, transparency by default is not privacy: anyone watching the list can reconstruct relationships. Third, identity built only for people: an agent, a model, or an industrial controller needs attested identity and a bounded permission — not only a wallet address.
Thesis: Zeqron is not “another token L1.” It is a semantic coordination protocol: parties negotiate and settle meaning-bearing agreements — contracts — with local actor state, optional privacy proofs, and post-quantum cryptography from genesis. Agents (ZAXON) and industrial assets (ZNA) are protocol concepts, not demos glued on at the end.
Walk one example — no invented customer, no figure. Two organizations agree on delivery of a composite batch. Their zone application presents a typed contract with context. Coordination validates it and, if the zone asks, attaches a transparent-setup zero-knowledge proof. Consensus orders the commitment on a DAG. Identity accompanies each party with post-quantum credentials. That path is the same object the design names in ContractState — Proposed → … → Anchored (with dispute and abort paths) — a pointer to the developer reading, without pasting code here. What the global record anchors is commitment order; publishing everyone’s balance is not required.
When this primer is enough: to fix problem, thesis, and the walked object. When to open another role: investors (diligence / gated pack); developers (four exhibits: ContractState / Phase / AgentIdentity / FinalityLevel); operators (enrollment still closed); enterprise (committee fit); legal (scope note); research (posture with labels). If site and brief diverge, the brief wins. ZQN is design vocabulary; this page does not offer an instrument.
Architecture · four layers
Public brief v0.2 §3 · design
- AApplicationSemantic Zones, domain modules, oracles
- BCoordinationContracts, STARK proofs, semantic bridges
- CConsensusQronGraph (DAG) + BFT-style committee
- DIdentityPQ DIDs, verifiable credentials, reputation
One job per layer. Conceptual flow: Applications → Coordination → Consensus; Identity authenticates throughout.
Walked case · composite batch
Design · no named customer · brief companion
- 01
Zone
Typed contract: batch delivery under a recipe (ZNA, concept).
- 02
Coordination
Proposed → Negotiating → Committed — or Rejected.
- 03
Proof
Optional STARK per zone. No public parameters here.
- 04
Anchor
Hash ordered on QronGraph. Anchored ≠ token transfer.
- 05
Agent
AgentIdentity bounds permission. Code change ⇒ new identity.
- 06
Fault
Aborted / Disputed exist. The object is not only the happy path.
Happy path and failure. Anchored is order on QronGraph, not a transfer. FRI parameters and amounts stay out (Appendix B).
Claim labels
Brief v0.2 · companion
design
Specified as protocol intent.
E.g. — PoSS = stake + observable service.
code
A type or family exists in the monorepo.
E.g. — ContractState, AgentIdentity, FinalityLevel.
unpublished
Withheld on purpose (NDA / phase).
E.g. — FRI, wire, yield, ports, jurisdiction.
If a citation drops the label, it is no longer this publication. Inventing a metric without a phase is a different class of error.
Principles the brief holds
Bilateral coordination
Work is negotiated between the parties first. The global DAG anchors hashes for order and finality. It is not a single world-account lock for every step of the agreement. That separates the protocol’s object — the meaning-bearing agreement — from a single transfer book.
Semantic typing
Relationships carry type and context inside zones. A number on an account does not describe what was agreed, between whom, or under which domain rule. Zones are where that rule lives.
Optional native privacy
STARK-class proofs with transparent setup are part of the design surface and are configured per zone. They are not a mixer bolted on later. Fine parameters (FRI, Poseidon, sizes) are not published on this site: deliberate omission from the brief.
PQ cryptography from genesis
FIPS 203 (ML-KEM) for key encapsulation, FIPS 204 (ML-DSA) for signatures, FIPS 205 (SLH-DSA) for long-lived signatures. The security plan is not “migrate later.” Classical curves in bridge experiments are not the long-term posture.
Agents and assets in the protocol
ZAXON binds agent identity to code digest, operator, and capability set: changing the code changes the identity. ZNA models industrial assets with genealogy, valid-composition recipes, and rules that prevent inventing mass. Both are concept-level in the public text; wire formats and market economics stay out.
Path friction (no invented customer)
Negotiation that does not close
Proposed → Negotiating can end in Rejected. The design contemplates “no”: not every agreement reaches Committing. That is useful friction, not a paper defect.
Execution that fails
Executing can go to Aborted; Disputed exists because the design admits conflict after commitment. Anchored is not “guaranteed success” — it is order on QronGraph when the path gets there.
Optional proof per zone
If the zone does not ask for a STARK, the path remains a valid typed agreement. If it asks and parameters are unpublished, this page does not invent a circuit: the friction is “under agreement / internal appendix.”
Agent over its bound
An AgentIdentity with bounded capabilities implies excess permission is not a feature. Changing the code changes the identity: the bound is part of the object, not a wallet afterthought.
Walked case — composite batch (design)
1 · Zone / Application
Two organizations in an industrial zone present a typed contract: delivery of a composite batch under a recipe (ZNA blueprint, concept). Status: design. No named customer, no SKU.
2 · Coordination
The contract enters Proposed → Negotiating. If both parties commit, Committing → Committed. If one refuses, Rejected and the path ends. Status: code (ContractState) · no public endpoint.
3 · Proof (optional)
If the zone requires privacy, a transparent-setup STARK is attached before or with the anchor. FRI parameters: unpublished here (brief Appendix B). Without the zone requirement, the typed agreement still stands.
4 · Consensus / anchor
The commitment hash is ordered on QronGraph (DAG, Mysticeti lineage adapted for PQ). Anchored = order, not a token transfer. FinalityLevel names claim states; it is not TPS.
5 · Identity / agent
If an agent runs a step, AgentIdentity (code_hash + operator_did + capabilities) bounds permission. Changing the code ⇒ new identity. Kill-switch and quotas: ZAXON concept; bond amounts unpublished.
6 · Useful failure
Executing → Aborted, or Disputed after Committed, are design paths. A committee that only reads the happy path is not reading the object.
What this primer is not
- A token offering or an issuance calendar.
- A node manual, or a minutes-to-first-app guide.
- A substitute for the Markdown brief: the brief is canonical.
- A claim of performance, finality, or availability of a production network.