cyb/prysm/proofs/proof.md

the proof layer of prysm

cyb is a soft3 interface. every value rendered on screen has a provenance — it is either cryptographically verified, asserted by an actor, or local to the neuron. prysm represents this provenance visually. the neuron always knows which values are proven and which are merely claimed

this page defines how proofs surface in the interface


the three states of truth

every value in cyb is in exactly one state:

state meaning source
proven cryptographic proof exists and verifies zheng proof, bbg opening, lens commitment
asserted claimed by an identified neuron but unverified signed cyberlink not yet finalized
local neuron's own state, never submitted local UI state, draft input, key material

the neuron sees the difference at a glance. proof is not metadata hidden behind a tooltip — it is the visual contract of the value itself


the verified glyph

verified state attaches a single visible mark to the value:

vector [verified, 2g, joy green]

the verified glyph (joy green ✓) appears adjacent to any proven value: a balance, a cyberank score, a neuron identity, a query result. it is the visual instance of "this is backed by a zheng proof"

the glyph is sized $2g$ inline, $4g$ in cards, $g/2$ as a corner dot on glass surfaces. consistent across atoms

atom verified rendering
text vector [verified, $2g$] appended after the text run
neuron vector [verified, $2g$] appended after the address
token vector [verified, $2g$] appended after symbol
field bright halo on the field surface, no glyph (value-per-point)

the asserted state

asserted values carry identity but not yet proof. a cyberlink in the mempool, a transaction signed but not finalized, a claim staked but not yet evaluated

asserted state renders with:

  • the neuron sigil of the asserter, $2g$ inline
  • no verified glyph
  • color follows emotion normally — no special tint

asserted is not a warning. it is the natural in-flight state of any submission to the cybergraph. it becomes proven when the next block finalizes


the local state

values that never left the neuron's device carry no signature, no proof. drafts in the commander, local UI preferences, key derivations. these render normally with no marker — the absence of identity is itself the signal

vector [verified, ...]   ← proven
vector [neuron-sigil, ...] ← asserted by a neuron
(nothing)                ← local

three states, three visual treatments. no ambiguity


proof inspection

every proven value is tappable. tap opens the proof inspector — an overlay molecule that displays:

field content
commitment bbg root hash at the block of proof
opening lens opening for the queried dimension
block finalization block height
algebra which strata tier produced the proof
size proof size in bytes
verify time local verification duration (always shown — keeps the neuron honest about cost)

the inspector is a prysm/display molecule with code atom rendering the proof hex. small. precise. the goal is not to make proofs pretty — it is to make them inspectable


the field as proof surface

the prysm/field atom is the densest proof surface in cyb: every point in the field's domain is a value, and every value can be backed by a proof. the field renders both:

  • the value (heatmap color, contour line, flow vector)
  • the proof density (a subtle saber overlay where coverage is dense vs sparse)

regions with full proof coverage glow with emotion at full intensity. regions with sparse coverage dim. the neuron sees the proof landscape

this matters most for the cyberank field: high-rank regions are also typically high-proof regions, because they have been queried often and proven often. low-rank regions are dim in both senses


staleness

proofs become stale. a proof generated at block $h$ may no longer reflect chain state at block $h + n$. prysm renders staleness as time-decay opacity on the verified glyph:

age (blocks since proof) verified glyph opacity
0 — current block 100%
1–10 90%
10–100 60%
> 100 30% — explicitly stale

the neuron sees a fading checkmark. they decide whether to re-query or accept staleness for performance. prysm never lies — stale proofs are visually distinct from fresh ones


the rule

emotion can soften. proof cannot lie.

if a value cannot be proven, do not render the verified glyph. if a proof has expired, fade it. if a query is in flight, show the loading state — never speculate forward to "probably proven"

prysm shows truth as it is, including its absence. the cybergraph is verifiable by design; the interface honors that design by never claiming verification it cannot demonstrate


ECS

  • Component: Provenance { Proven { proof_cid, block_height } | Asserted { neuron_id } | Local }
  • Component: ProofFreshness { block_age }
  • System: ProvenanceSystem — reads chain state, writes Provenance to entities
  • System: VerifyGlyphSystem — for Proven entities, attaches verified vector child with freshness-driven opacity
  • System: ProofInspectorSystem — opens inspector overlay molecule on tap of verified glyph

the cybergraph is a graph of proofs. prysm is the lens that shows them

Homonyms

prysm/proof
cybics/crystal/proof
zheng proofs cover every cyber/signal atomically — cyberlinks, conviction UTXOs, and cyber/impulse in one verification. see cybergraph/reference/proof.
cybergraph/reference/proof
soft3/strata/proof
proof
soft3/mudra/src/proof
proof
neural/inf/specs/proof
inf proof how an inf result certifies itself against a bbg root. a query result, or the set a mutation appends, carries a zheng proof that it follows from the committed graph state. a verifier holding only the 32-byte root checks it in constant time. what a proof asserts for a read, the proof binds…
soft3/cybergraph/specs/proof
proof the σ field of a signal: a single zheng proof. it is the signal's **evidence** — one proof, covering the whole batch atomically, that lets anyone holding only the root accept the signal without re-executing it and without trusting the sender. what it is σ is produced by the neuron at sealing,…

Graph