RISK. INVESTIGATIONS. PROTOCOL INTELLIGENCE.Meet your intelligence toolkit ↗
Governance & treasury forensics

Was every payment
actually approved?

Reconcile each treasury movement against the proposal that authorised it, examine who really controls the multisig, and see where voting power concentrates, with every finding ranked by severity and the DAO graded A to F.

Not saved yet. Keep the report to add it to the case, or discard it. Deleted automatically in 72h if undecided.Keep reportDiscard
C65/100
Governance & treasury audit · Ethereum · 2026-10-14 10:05Example DAO
CRITICAL 0HIGH 2MEDIUM 2LOW 1
  • HIGHOne voter decided "Market-maker loan"
  • HIGHSafe threshold is 2 of 5
  • MEDIUMSnapshot proposals have no quorum
Safe 0x5e2b…9d41 · 2 of 5 · 214 transactions Treasury outflows reconciled with proposals
DateAmountToApproval
2026-09-30250,000.00 USDC0x1c07…e4a2Grant to Core Devs
2026-10-031,200,000.00 USDC0x8f5d…20b1Market-maker loan
2026-10-093,400.00 ETH0x44ae…7c13No matching proposal

Automated findings are review leads for an investigator, not legal conclusions.

15largest treasury outflows matched to proposals
25recent Snapshot proposals read
A–Fgrade, with every finding behind it
0findings treated as proof: each one is a lead
Treasury reconciliation

Every large outflow,
checked against its approval.

Each of the largest treasury outflows is matched to a closed Snapshot proposal that names the recipient, or to an on-chain proposal executed in the same transaction. Hover a row to see why.

Treasury outflowValueRecipientTimingApproved byResult
Grant to Core DevsA closed Snapshot proposal names the recipient and closed two days before the payment.250,000 USDCNamedClosed 2d beforeSnapshotMatched
Market-maker loanMatched, but one voter cast 58% of the votes on it: a separate finding.1,200,000 USDCNamedClosed 1d beforeSnapshotMatched
Upgrade paymentPaid in the same transaction that executed on-chain proposal 17.80,000 USDCNot neededSame txOn-chainMatched
Treasury rebalanceNo recent proposal names this recipient. It may be routine or unauthorised; ask the DAO for the approval.3,400 ETHNot named—None foundNo match

Unmatched outflows may be routine (payroll, swaps, grants approved elsewhere) or unauthorised; five or more of them is a high finding.

AreaWhat it readsWhat it flags
Snapshot settingsAdmins, voting period and delay, quorum, who can proposeVoting window under 24h; no quorum; one admin
VotesThe five most recent closed proposals, voter by voterOne voter over half; three or fewer hold a majority
Safe multisigSigners, threshold, modules, guard, signer account types1 signature; under half; modules that skip signing
Governor & timelockVoting delay and period, proposal threshold, timelock delay and adminNo timelock; delay under 24h; single-wallet admin
Governance tokenTop holders, leaving out DAO-owned and locked walletsOne holder over a third; top ten over 80%
TreasuryOutflows by recipient; the largest matched to proposalsOutflows no proposal approved
Six sources

Six places it reads,
each with its own checks.

The audit reads Snapshot, the Safe, Governor and timelock contracts, the governance token and the treasury’s transfer history directly. Every address it meets is screened against public scam blocklists.

  • The grade summarises the findings; it is not a fraud score.
  • Records that could not be read are listed as limits of the audit; they are never counted as misconduct.
Published thresholds

The lines we flag,
and where they sit.

Largest voter’s share of a proposalObserved: 58%
50%

Above half, one voter decided the proposal: a high finding. Three or fewer voters holding a majority is medium.

Voting windowObserved: 14 hours
24h3 days

A fixed window under 24 hours is high; under three days is low. Timelock delays under 24 hours are flagged too.

Multisig threshold vs signersObserved: 2 of 5 · 40%
50%

One signature is critical; fewer than half the signers is high; a Safe with only two signers is medium.

Findings with next steps

Every flag tells you
what evidence to get.

A finding is a lead, so each one states what was observed and exactly what to do next. When you keep the report, the findings and the addresses to follow join the case.

A–F when records were readN/A when none were

A DAO with nothing to read gets no grade, so a thin record never passes as a clean one.

  1. 5 of the 15 largest treasury outflows match no proposalHigh

    Next: Ask the DAO for the approval behind each transfer. Unmatched outflows are the first place to look for misappropriation.

  2. One voter decided “Market-maker loan” (58% of votes)High

    Next: Trace how this voter got its voting power (token purchase, delegation, flash-borrow) shortly before the snapshot block.

  3. Safe has 1 enabled module that can bypass signaturesHigh

    Next: Audit each module contract and list the transactions it executed from the Safe.

  4. Snapshot proposals have no quorumMedium

    Next: Look at the vote count of every proposal that authorised spending.

Match every outflow to its vote.See which treasury movements governance approved, and which it never did.
Questions

For protocols
and token holders.

Something else? Ask our team →

  1. 1

    Who is it for?

    Protocol and DAO teams, token holders and investors, and investigators looking into a treasury or governance incident.

  2. 2

    What do I need to start?

    Any one of: a Snapshot space or proposal link, a Safe or treasury address, or a Governor or timelock address. The audit fetches the rest itself and writes nothing until you keep it.

  3. 3

    How is the grade worked out?

    Each review starts at 100 and loses 25 points per critical, 12 per high, 5 per medium and 1 per low finding; any critical finding caps it at 39. A is 90 or more, then B from 75, C from 60, D from 40 and F below. The findings are listed with it, so you can see exactly what moved it.

  4. 4

    Do missing records count against the DAO?

    No. Anything that could not be read is listed under the limits of the audit. If nothing could be read at all, the DAO gets N/A instead of a grade.

  5. 5

    How are outflows matched to proposals?

    The largest outflows (five per asset, up to 15) are matched to closed Snapshot proposals that name the recipient and closed in the 120 days before the payment, or to on-chain proposals executed in the same transaction. Anything left over is shown as “No matching proposal”.

  6. 6

    What does the output look like?

    A graded report: findings by severity, each with the next step to take, the Safe and its signers, Snapshot proposals and treasury outflows matched to the proposals that approved them. It joins the case when you keep it.

Reconcile your treasury
before someone else does.

Bring a DAO or a treasury incident. We’ll walk through the reconciliation, the grade and the findings with you.