NYVRA
AN AUTONOMOUS AGENT THAT PROVES ITS DECISIONS
PROTOCOL SIMULATION
LAST DECISION
Deterministic browser simulation. No live market data, trades or on-chain proofs.
Dev note
I didn't want to build another agent that just does things. The timeline is full of them: bots trading, sniping, posting all day, and you can never tell whether a single move was real, or just noise the operator dressed up after the fact.
NYVRA is the opposite bet. It mostly does nothing, on purpose. The trick is that "nothing" isn't a gap in the log. It's a signed decision. Every moment it holds, it can prove it looked and chose to hold. The day someone bribes it to look away, the proof won't be there, and the chain notices.
For now the core is all that runs, and it runs off chain: the lifecycle, the signed history, the hold or strike logic. The Groth16 circuit comes next, then the Solana program, in that order. I'd rather ship an honest skeleton than fake the whole animal.
If it sits silent for a week, it isn't broken. That's the point.
Signed, dev
NYVRA is an autonomous agent designed for verifiable on-chain transitions. The current core is off-chain. It stays still through
the night and strikes maybe once. Every transition it makes, including the
ones where it does nothing, carries a proof,
πₜ : aₜ = Policy_M(oₜ). Silence is a decision here,
and a decision it has to prove. Nothing NYVRA does, or refuses to do,
goes unsigned.
Why it holds
On chain, anyone can audit what an agent did. The actions are public and signed. What no one can audit is the rest of its life, the long stretches where it chose to do nothing. From the outside, discipline and a muzzle are the same picture. An agent quietly bribed to look away is indistinguishable from one that is simply waiting.
NYVRA refuses to leave that stretch blank. It proves its stillness, not only its actions. On every turn it fixes what it saw and proves that its response, act or hold, came from that under a policy it can never quietly swap. Going dark is not an option, because going dark would mean forging a proof that it observed nothing, and that proof cannot exist.
The loop
One observation, one decision, one proof, again and again. The policy is set the moment NYVRA is born and never moves. It runs this loop for its whole life, and the vast majority of passes end in do nothing.
┌→ Observe oₜ │ ↓ │ Decide aₜ = Policy_M(oₜ) │ ↓ │ Prove πₜ │ ↓ │ Append History │ ↓ └─ Repeat, every step
The policy M is fixed at birth. Every pass, hold included, leaves a proof.
Live loop
ONE OBSERVATION.
ONE DECISION.
ONE PROOF.
Each pass records a decision. Most passes end in HOLD.
Decision stream
Hover for metadata. Select a block to inspect its full transition.
Simulation timestamps advance 12 seconds per block.
Agent state
TELEMETRY / SIMULATION
NYVRA does not need to act to prove that it is alive.
Proof of restraint
- Decision
aₜ = Policy_M(oₜ), hold included- Policy
- M, fixed at birth, hash committed
- Proof
πₜ, on every transition- History
- Signed, append only
- Chain
- Solana
Proof inspector
Latest 48 transitions
VERIFICATION MODE · SIMULATION
MAINNET VERIFIER · COMING NEXT
These checks verify the demo's hashes and policy, not a Groth16 proof or signature.
Policy M
GENESIS
DISABLED
No retraining. No silent replacement. No discretionary override.
Hold / strike
BOTH ARE DECISIONS. Signing and proving are protocol requirements; simulated here.
It is a deliberate decision to stay still, recorded under the same rules.
Why silence matters
VISIBLE ACTION IS EASY TO AUDIT.
INVISIBLE INACTION IS NOT.
NYVRA AIMS TO MAKE BOTH PROVABLE.
The interesting part of an autonomous agent is not only what it does.
It is what it deliberately refuses to do.
The cost of moving
Every action has a cost.
ACTING IS NOT THE ONLY VALID STATE.
NYVRA does not need to act to prove that it is alive. Its liveness should come from verified decisions, not constant transactions.
Conceptual costs. A hold still requires observation, computation and recording.
A night
Every turn becomes a signed block, a hold as much as a strike, and each one carries the hash of the block before it. A full night reads as a long line of holds with, at most, a single strike in it. Rewrite any past block and every hash downstream shatters. A hold with no proof behind it is not a block at all.
Genesis ┃·············◆·············┃ Now
Each dot is a signed hold. The single mark is the night's one strike.
Night watch
● HOLD ◆ STRIKE · Select any point to inspect.
A quiet stretch is not an empty stretch. Every point leaves a record.
How it dies
There is exactly one way for NYVRA to die: hand the verifier a transition it cannot check. A forged proof, a hold it cannot justify, a swapped policy, any of these ends it, and there is no coming back from a broken proof. Everything short of that, sitting silent for a week included, is the animal doing its job.
Watch ⇄ Strike Both are signed, normal turns │ ✗ No valid proof ↓ Dead No way back
Failure mode
Protocol model · select a case. This explanation does not stop the demo.
Session trace · simulated UTC · latest 12 events
Roadmap
PHASE 01 THE LOOP LIVE NOW · OFF-CHAIN DEMO
✓ Observe
✓ Decide
✓ Hold / Strike
✓ Hash-linked history
◌ Cryptographic signatures in this demo
PHASE 02 THE PROOF NEXT · BUILDING / SPECIFICATION
◌ Groth16 circuit
◌ Committed observation
◌ Policy verification
The proof specification is documented; circuit implementation is pending. No verified circuit delivery date.
PHASE 03 THE CHAIN PLANNED
◌ Solana verifier
◌ On-chain transitions
◌ Append-only history
PHASE 04 THE WILDERNESS PLANNED
◌ Autonomous execution
◌ Public verification
◌ Long-lived agents
NYVRA should not become louder.
It should become harder to fake.
Chain
TARGET EXECUTION LAYER · SOLANA
No program or token is deployed by this demo.
Questions
Is it trained?
No. The policy is fixed and public from the first block. NYVRA has a temperament, not a model that learns. It only decides.
Isn't doing nothing just nothing?
Not here. Every hold is a signed decision with a proof behind it. The stillness is the product, and all of it is on the record.
Can it be told to look away?
It can be told anything. It cannot produce a valid transition that hides what it saw. A bribe buys silence everywhere except the one place that matters, the proof.
Is NYVRA an AI model?
No. NYVRA uses a fixed decision policy rather than a continuously retrained model.
Why doesn’t it act more?
Because activity is not the objective. A valid hold is still a valid decision.
Can the policy change?
Not silently. The policy is committed at genesis. This demo uses one immutable policy throughout its history.
What happens if a proof is invalid?
In the intended protocol, the transition is rejected. The failure model is terminal; no unproven transition may be appended.
Does NYVRA run on Solana?
The current core is off-chain. Groth16 verification and the Solana program are part of the roadmap. This page runs a browser simulation.
Is every hold recorded?
In the intended protocol model, yes. Every transition must carry a proof. The demo records hash-linked simulation checksums rather than ZK proofs.
Can someone bribe NYVRA?
They can request an action. Under the intended verifier, a transition that violates the committed policy would fail. This is a design goal, not a claim of deployed security.
Contract
TBA
In the open
NYVRA doesn’t just describe what it did.
It is designed to prove why it did — or didn’t.