One Zero-Gas Vote Moved a $180,000 Treasury Decision

One green Sign button, one wallet popup, zero dollars of gas. March 9, 2026, 6:55 am, standing at my kitchen counter with cereal goin' soggy while I read a governance proposal on my phone, which is a sentence my younger self would have found deranged. The proposal was number 42 in a slight indexing DAO I have held tokens in since 2023. The ask: $180,000 from the treasury for one full year of a security developer. My thumb hovered. Ninety seconds, maybe more.

That treasury is real money with a public address: $1.9 million in a Gnosis Safe, a 4-of-7 multisig whose every move anyone can read on Etherscan, cuz that is what treasuries mean on-chain. The DAO itself is modest, a data-indexing project with maybe 400 active voters and a Discord that goes quiet for days at a stretch. My entire stake is 1,400 tokens outta approximately 155,000 voting weight. About 0.9 percent. Historically I assumed that meant nothin', and proposal 42 was about to test the assumption in both directions.

The pitch was solid on its face: $15,000 a month for twelve months to a developer known in the community as ren, funded to audit access controls, dependency updates, and the indexer's permission model before the Q4 mainnet changes. What the proposal did not contain was any cliff, any vesting, or any reporting obligation past a vague monthly update. A person could take the $180,000, vanish in month five, and every line of the proposal would still have been honored. I read that gap four times before I believed it. No cliff. No vesting. The gap gets fixed on March 12, and the fix is three lines long.

the treasury, on Etherscan

Snapshot voting costs nothin' because nothin' goes on-chain; the signature lives off-chain and the result gets executed by the multisig later. That felt fake to me the first three times I voted, so I did my inspecting elsewhere: every Safe transaction, every payout, every treasury balance sits readable on Etherscan, and I keep the 4-of-7 signer list bookmarked. Gas was $23 a vote back in 2021, the one time I voted on-chain, and I never did it again. Free voting made me suspicious. Reading the wallet made me useful.

ren, the 2,400-word post

Ren's own forum post, linked from the proposal, ran 2,400 words and included two prior security reviews he had done for free, a scope document, and a burn-down schedule by month. I dug into it on the train and came away believing he was real, which made the missing cliff stranger, not less peculiar. The proposal itself compressed all that care into eight paragraphs and a budget table. Twelve months at $15,000, one line for $4,800 of audit tooling, total ask $180,000 against a $1.9 million treasury, call it 9.5 percent of everythin' the DAO holds. I chewed on the ratio for two days.

the vesting question I left at 1 a.m.

At 1:05 am on March 11 I posted a three-line comment on the proposal thread instead of voting. It asked one question: what stops ren from taking months one thru five and stopping. I suggested a six-month cliff with a quarterly report requirement, and I said plainly that I would vote against anything without one, and I meant it in writing. The threat was almost funny. Two people holding 6,000 combined tokens quoted my comment within nine hours.

three lines, changed March 12

Ren amended the proposal on March 12, sixteen hours after the comments began, and the amendment contained three changes: a six-month cliff, quarterly reports with published commits, and a 30-day cure clause. His note ran two sentences and included the phrase fair point, which is governance for I folded. Two sentences. The forum thread went from 11 comments to 63 in two days. My original three lines got quoted, screenshotted, argued with, and once misattributed to the project lead. Three sentences, zero dollars, one clause that outlives every vote cast.

the quorum math

Lumpy, in other words. Weight distribution in this DAO is lumpy enough to matter, since the top five wallets hold 31,000 combined, delegations drift around by the week, and a typical proposal draws 90,000 to 130,000 of the 155,000 total weight. My 1,400 was 0.9 percent of supply and maybe 1.1 percent of what actually voted on 42. Modest, but not decorative, and the margin came in at 1,900. Numbers that size are why I no longer skip a breakfast vote.

signing cost nothing, twice

I voted For on March 14 with a signature that cost $0.00 and about eleven seconds of wallet popup. The Snapshot receipt page shows my 1,400 weight sitting among the 407 wallets that voted, everythin' timestamped, everything public, nothin' on-chain, and nothin' about it on trial in any court. Against finished at 61,200. For finished at 63,100. My 1,400 sat inside a 1,900-token margin, and whether it was decisive is unanswerable, and whether it was part of the margin is arithmetic.

April 2, the payout

The Safe executed on April 2, four transactions, $45,000 on day one covering the first quarter plus tooling, and I watched the outgoing transfer hit Etherscan within a minute of the Discord announcement. The remaining quarters release only after each report clears the cliff and the review. Ren published his first on time: 14 pages, three findings, one of 'em an access-control bug he rated high severity and fixed before anything touched it. His estimate of what that bug could have cost in a bad quarter: more than the entire grant. No drama attended any of it. Next time proposal 43 lands on my phone at breakfast, the cereal will still lose. The green button costs nothing. That was the whole discovery.