A wallet-verifiable strategy evidence network for indexing versioned trading systems against exact Solana token mints.
Concept paper
v0.1
August 7, 2026
Proof means attributable evidence—not future profit. This framework does not promise returns, guarantee recovery, grant custody, or authorize a transaction. Every wallet remains independently controlled.
Abstract
Proof before trust
A participant commits to a versioned PIT Card before entry, retains custody, completes or abandons the position on-chain, and allows Trinity to measure the outcome after observable costs. The result is a durable receipt of what was intended, what occurred, and how it was classified.
The index
Strategies × exact token mints
Each cell accumulates qualified observations across independent wallet clusters, time periods, liquidity conditions, and volatility regimes. Trade Wars is the experience layer. Proof-of-Strategy is the evidence system beneath it.
01 · The problem
Transactions alone do not prove a strategy
Trading performance is often communicated through screenshots, selective leaderboards, unverifiable backtests, or claims made after the result is known. Even an authentic on-chain round trip proves only that a wallet traded. It does not identify the strategy that motivated the trade or establish that the result can be reproduced.
What is missing
A commitment made before entry
An immutable strategy version
Exact asset identity by mint
Complete win, loss, open, and abandoned outcomes
What must be measured
Fees, slippage, holding time, and drawdown
Liquidity and volatility at entry
Wallet independence and concentration
Evidence recency and evaluator version
02 · Core evidence unit
The voyage receipt
wallet or wallet cluster
+ strategy version
+ pre-entry PIT Card commitment
+ exact Solana token mint
+ entry conditions
+ finalized entry and exit records
+ net outcome after observable costs
+ evidence classification
More wallets do not automatically produce better evidence. Evidence improves only when observations are complete, economically meaningful, independently attributable, and measured under declared rules.
03 · Meaning of proof
Attribution has a boundary
It can establish
A wallet controlled the signing key
A PIT Card existed before entry
The card named an exact strategy version and mint
Referenced transactions finalized
A declared evaluator produced the classification
It cannot establish
That a strategy will work again
That different addresses mean different owners
That liquidity or token value will persist
That averaging prevents permanent loss
That a strategy is safe or suitable for everyone
Proof-of-Strategy is proof of attributable evidence, not proof of future profit.
04 · Voyage lifecycle
Commit first. Measure everything.
DiscoverCommitEnterResolve
Discover and commit
Trinity checks exact mint identity, route, liquidity, volume, concentration, cooldown, and wallet capacity. Before entry, it produces a wallet-bound PIT Card with the strategy version, mint, limits, expiration, and canonical commitment hash.
Observe and resolve
Finalized entries, adds, exits, costs, adverse movement, and deviations remain attached to the voyage. A voyage resolves as profitable, losing, expired, abandoned, invalid, or still open. None of those states may disappear merely because it is unfavorable.
05 · Outcome accounting
Purchasing-power change after friction
net outcome
= realized exit value
− matched acquisition basis
− network fees
− protocol and routing fees
− observable slippage
USDC, SOL, or another declared numeraire may be used, but the numeraire remains attached to the record. If complete basis cannot be established, the result stays partial or unverified. Unknown data never becomes a favorable assumption.
06 · Qualification
One hundred observations—not one hundred selected winners
A future gate may require extensive independent confirmation before Trinity presents a strategy–token combination. Counting only profitable trades would create survivorship bias and corrupt the index. Every confirmed precommitted entry must retain its final state.
Every qualified observation
Was committed before entry
Names the exact mint and network
Uses finalized, economically matched records
Includes costs under a declared method
Passes anti-dust and sampling-window rules
Illustrative community gate
100 completed qualified observations
40 estimated independent controller clusters
At least 30 calendar days
Multiple volatility regimes
Positive median net outcome and conservative confidence bound
These thresholds are a research starting point, not a permanent assertion of statistical sufficiency.
07 · Independence
One address is not necessarily one participant
A controller can create many wallets, repeat dust trades, or route circular activity to manufacture confirmation. Trinity should estimate controller clusters using explainable public signals such as funding paths, repeated counterparties, timing correlation, transaction structure, and direct self-routing.
SamplingOne weighted vote per cluster
ConcentrationPer-cluster contribution cap
DustExcluded
UncertaintyDisclosed, not hidden
Clustering remains a probabilistic interpretation and must be labeled accordingly.
08 · Confidence states
Evidence can mature—and become stale
UnobservedExperimentalObservedQualified
STALEEvidence no longer reflects current conditions
SUSPENDEDIntegrity, identity, liquidity, or manipulation concern
VERSIONEDEvery score names its rules and evaluator
Qualification changes the amount of verified historical support for presenting a setup. It never turns history into a guarantee.
09 · NFTs
Identity and eligibility—not authority
NFTs may identify ships, strategy templates, versions, fleet links, achievements, helper records, or presentation access. Ownership does not grant custody, signing power, delegated execution, promised rewards, or proof of strategy effectiveness.
10 · Reputation
The chain proves the event; rules supply meaning
A transfer can support a voluntary helper credential, but the transaction alone does not prove motive. Interpretation must come from a signed claim or declared rule, remain reviewable, and never create custody over another wallet.
11 · Cooperative boundary
Share evidence, not market impact
Fleets may collaborate through analysis, attestations, and independently produced evidence while every wallet retains custody and makes its own decision. Trinity must not organize synchronized buying, artificial volume, wash trading, price support, or coordinated exits. Activity designed to move a market or help another participant exit does not qualify as cooperative evidence.
The network’s value comes from independent reproduction—not collective price pressure.
Raw transactions are durable public facts. Normalization, clustering, regime classification, and scoring are interpretations. Trinity versions those interpretations so the system can learn without pretending that a new model was always the original result.
13 · Self-custody
Every wallet remains its own vessel
Keys and recovery phrases never enter Trinity
Message signatures do not move funds
Every transaction remains separately reviewable
Fleet links are analytical, not pooled custody
Revocation ends future access without rewriting history
14 · Fail closed
Unknown stays unknown
Unverified token identity
Wallet or commitment mismatch
Unfinalized transaction status
Unknown basis or cost accounting
Stale liquidity, volume, or evaluator version
15 · Additive compatibility
Learn without rewriting the past
Prior strategy versions and voyage records remain inspectable. New learning may publish a successor strategy, tighten a gate, suspend an evidence cell, introduce a new evaluator, or add a presentation layer. It must not silently change the rules attached to an old commitment or erase an unfavorable outcome.
16 · Development path
Build evidence quality before network scale
Personal evidence ledger: validate founder-wallet commitments, exact-mint identity, complete outcomes, and repeatable evaluation.
Founding cohort: test a bounded group of self-custodied wallets, independence weighting, and revocation.
Public evidence explorer: publish read-only cells, methods, confidence states, and limitations.
Community confidence gates: let qualified evidence influence which PIT Cards are presented while owners retain every decision.
Open research layer: compare authenticated strategies under shared schemas without endorsing every author or collection.
17 · Risks
Evidence does not eliminate market loss
Proof-of-Strategy remains exposed to regime change, illiquidity, incomplete cost data, clustering errors, malicious participants, smart-contract failures, token collapse, evaluator bugs, and human misinterpretation. Averaging can lower the price required for recovery while increasing capital exposed to a declining asset. No amount of averaging guarantees recovery, and an asset may approach zero or become impossible to exit.
Category definition
A wallet-verifiable strategy evidence network
Proof-of-Strategy is a framework in which participants precommit to versioned trading rules, retain custody, contribute complete finalized outcomes, and allow those outcomes to be evaluated across exact assets and market regimes.
Trade Wars is Trinity’s interactive expression of the framework: wallets become vessels, strategies become rigs and PIT Cards, completed trades become voyages, and the blockchain preserves the receipts.
Proof before trust.
Complete outcomes before leaderboards.
Independent evidence before confidence.
Self-custody before participation.