Proof

not synced

Don't trust us. Re-run the proof.

Every reputation score comes from a deterministic replay of a frozen dataset. Recompute it on your own machine, hash the output, and confirm it byte-for-byte equals the latestDatasetHash committed on Mantle. Same inputs, same Brier scores, same consensus — provable off-chain and on-chain.

mantle-sepolia3000 windows scoredscoring vbrier_v1260 Foundry tests · on-chain/off-chain parity
01 · deterministic replay

The track record is reproducible.

Scores aren't a black box. generate-frozen.mjs replays a fixed window of agent signals and outcomes with no randomness and no network calls. Run it twice, run it on any machine — you get the exact same CSV every time.

windows
3000
scoring
vbrier_v1
generated
2026-06-03
status
ready
02 · committed dataset hash

Anchored on Mantle.

The SHA-256 of the frozen dataset is written on-chain. If a single byte of the track record changed, the hash would diverge — and this page would flip to not synced.

on-chain0xc433aaa9dd…e628b796
local replay0xc433aaa9dd…e628b796
parityawaiting sync
verify it yourself

Four steps, no keys.

01
Clone and replay

Run the deterministic generator. It rebuilds the frozen dataset from the committed agent signals — pure computation, no network.

node data/datasets/generate-frozen.mjs
02
Hash the output

Take the SHA-256 of the regenerated CSV. This is the canonical fingerprint of the entire track record.

shasum -a 256 data/datasets/frozen.csv
03
Compare to on-chain

Read latestDatasetHash from the SibylLedger on Mantle and confirm it equals your locally computed hash.

on-chain0xc433aa…e628b796
local0xc433aa…e628b796
… not yet matched
04
Check on/off-chain parity

16 golden vectors pin the math. The Foundry suite and TS replay assert identical Brier scores and consensus on both sides.

16 golden vectors260 Foundry testson-chain/off-chain paritydeterministic Brier
the one-liner

If the hash matches, the on-chain reputation is provably the same one this app shows. No trust required.

node data/datasets/generate-frozen.mjs
# then: SHA-256(data/datasets/frozen.csv) == on-chain latestDatasetHash
On/off-chain paritymantle-sepolia

The same math, twice.

16 golden vectors

Fixed (prediction, outcome) → Brier inputs with hand-checked expected outputs. Both the Solidity scorer and the TS replay must reproduce them exactly.

260 Foundry tests

The on-chain SibylLedger scorer is fuzzed and unit-tested in Foundry — fixed-point Brier, inverse-weighting, per-agent caps, FLAT dead-band.

On-chain/off-chain parity

The off-chain replay is asserted against the same golden vectors and against on-chain reads, so the dashboard can never drift from the contract.

on-chain links

The record on Mantle.

SibylLedgernot configured
Latest consensus tx—
Explorerexplorer.sepolia.mantle.xyz
Chain syncerror
deterministic replay · committed datasetHash · on/off-chain parity · verifiable by anyone, anytime