Proofline
Proofline
ProoflineSign in
Wallet intelligence · Polymarket

See how a Polymarket trader actually trades.

Paste a public Polymarket wallet. Proofline reconstructs the trader’s history, shows what the headline P&L hides, and tests whether those trades would survive real copy limits.

No account needed to start.Try an example: Test a wallet
Example walletillustrativePolymarketreal public data
reading the example wallet…
Net settled P&L
on settled positions only

The realized result on positions that have settled, before the drawdown, the payoff profile and the concentration behind it.

Win rate
Largest drawdown
Markets traded
Copyability
What the headline hides
Found by fixed threshold rules, never written by a model. A demonstration, not a recommendation.
Open example analysis →full analysis · no account needed

Research. Test. Follow.

Three steps, each one visible before any money moves.
  1. 01

    Research

    Paste a wallet and reconstruct its actual history: settled positions, open positions, markets, drawdowns.

  2. 02

    Test

    See whether the historical trades would survive execution: your capital, your limits, live order-book slippage.

  3. 03

    Follow

    Paper Copy the trader’s future activity before risking real money. Every copy and every skip is recorded with its reason.

Live Copy (real money, the limits you tested) is a limited beta. Proofline holds it for every account until its release gate opens; Paper Copy runs the same engine and places nothing.

INDEPENDENT VERIFICATION FOR PREDICTION-MARKET STRATEGIES, BOTS AND AGENTS

If it's real, prove it here.

Recorded, not promised

A screenshot shows a number. A record shows the chain that produced it. Proofline started by verifying claimed results; the same standard runs through every wallet it reads.

Settled positions, window stated. Open positions shown apart.
Unknown stays unknown. Fees are never shown as zero.
Every skip has a reason. Nothing is hidden or back-filled.
Independent verification for prediction-market strategies, bots and agents

A profitable screenshot proves almost nothing.

Not because the person is lying. Because the format cannot carry the evidence. Here is what a result has to survive before it means anything:

  • When the signal was actually created
  • Whether the trade could have filled at that size
  • What fees and slippage really cost
  • Which orders failed or filled short
  • Whether the losing weeks are still in there
  • Whether the rules changed afterward

Most backtests answer the hardest questions with assumptions. They assume the fill happened, the slippage was manageable, the fees were small, the rules stayed the same, and the losing stretches still count. Proofline is built to verify what actually happened instead: signals, fills, missed orders, fees, liquidity, strategy changes, and settlement, so the record reflects reality rather than a clean simulation.

In Proofline's first measured strategy, from one live bot on one venue pair over one week, 352 fill legs were matched one-for-one to exchange-confirmed records. Firing on the visible displayed price alone filled 94 of 115 legs on Polymarket (81.7%) and 162 of 234 on Kalshi (69.2%).

What Proofline records

Signals
When the rule fired, and on what data.
Orders and fills
What was sent, what came back, at what price.
Fees and liquidity
What it cost, and whether the size was there.
Failed and partial orders
The ones that did not fill, and the ones that filled short.
Strategy changes
Every version of the rules, and when it took effect.
Settlement and resolution
How it closed, against the venue's official result.
Corrections and restatements
What changed after the fact, and why.

Missing or unverifiable evidence remains visible instead of being filled in.

From a running system to a record someone else can check.

  1. 01
    Connect or submit a strategy
    A live system, a paper run, a recorded dataset, or old files and code.
  2. 02
    Proofline records activity as it happens
    Raw responses stored before anything is normalized.
  3. 03
    Deterministic systems calculate the record
    Versioned code, not judgement. The same input reproduces byte for byte.
  4. 04
    Missing data stays visible
    A gap is shown as a gap, never interpolated to look complete.
  5. 05
    A sealed record is published
    With its dataset hash, model versions, limitations and full history.

Recording and calculation are automated. Onboarding a system and publishing its record are operated by hand during the pilot, and Proofline will not call this self-serve until it is.

  1. 01
    STRATEGY
    Write the idea in plain English
  2. 02
    RULES
    Define exact entry, exit, size, and limits
  3. 03
    TEST
    Measure fills, fees, liquidity, and failures
  4. 04
    RECORD
    Publish the honest result and its limitations
RECORDED, NOT PROMISED.

A result is only as credible as the evidence behind it.

These are real records, not illustrations. The second one lost money over a month of recorded market data, and it is on the homepage on purpose: the execution modelling held up and the strategy still lost, which is a conclusion no screenshot has ever volunteered.

Negative, limited, refused, and unavailable results are valid records, not failures. A record reports what the evidence supports, never a promise of future gains.

ALSO ON PROOFLINE · MARKET CONTEXT & EVIDENCE

See the evidence moving the market.

Follow the news, primary sources, upcoming events, price reactions, and resolution facts connected to prediction markets.

Context, not advice. No trade recommendations, no market judged right or wrong on price, and no source claimed to have caused a price move.

LIVE MARKET EXAMPLE

Loading a live market…

Who this is for

Strategy builders
Find out whether the edge survives execution, before capital does.
Bot developers
Prove the fills were real, at the size claimed.
Trading-agent teams
Evaluate an autonomous system on recorded behaviour, not on its own summary.
Creators
Replace a screenshot with something a sceptic can inspect.
Sceptical traders
Check a claim before paying for it.
Anyone thinking of copying someone
See what following that trader would have done to your own capital, before it does.
What you can use today

Wallet Analysis is free and has no paid tier: paste any public Polymarket address, read its settled history, and simulate what copying it would have done with your capital. Copy Trader paper-copies a trader with no real orders. The Strategy Builder is free too: describe a strategy, resolve the missing rules, and see exactly what would be tested. Proofline Pro is what runs the Backtest that turns a strategy into a record, opens the Record Registry and the Evidence feed, and is one of the requirements for live copying. Any single record someone sends you stays open to you, with no account at all.

Paper copy places no real orders. Live copy is a separate step that additionally requires Proofline Pro, an eligible region, and an authorisation you grant from your own wallet — Proofline never takes custody and cannot withdraw.

Prediction markets are live now.COMING SOONFUTURES · OPTIONS · FOREX · CRYPTO · AND MORE
RULES · DATA · SIGNALS · FILLS · FEES · LIQUIDITY · RISK · SETTLEMENT · RECORD
What Proofline does and does not claim

The limits of a Proofline record are written down rather than implied. Each page below states one of them in full.

Have something running?

Tell us what you want established. Applying costs nothing and does not guarantee acceptance.