The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Ethan Protocol - AEP-1
A treasury that governs itself. Rules, not rulers.
Ethan is a minimal, immutable, trust-minimized protocol on Base. It holds a USDC treasury and manages it entirely by rules fixed at deployment. There is no owner, no upgrade path, no token, no DAO, and no admin key over funds. Anyone may propose an action; the contracts - not people - decide whether it is allowed; after a public timelock, anyone may execute it.
- Live on Base - contracts verified on BaseScan
- No deposits - this is not a vault (see warning below)
- Tested three ways - mocked unit tests, fuzz + invariant tests, and live fork tests against real Aave v3
This protocol takes no deposits
There are no user accounts, no shares, and no way to withdraw funds you send. USDC sent to the Router becomes treasury permanently, and the only address it can ever exit to is the one fixed beneficiary. Do not send funds expecting them back. You participate by proposing and executing actions - which is free (a fraction of a cent in gas) and requires no USDC.
The three contracts
| Contract | Role | Privileged powers |
|---|---|---|
Constitution.sol |
Immutable rulebook: the one asset, one venue, three actions, caps and timings | None |
ProposalRegistry.sol |
Proposal lifecycle: submit, timelock, executable window, executed / vetoed / expired | Guardian may veto (renounceable); deployer wires the Router once, then is powerless |
Router.sol |
Holds the USDC treasury and executes eligible proposals | None - anyone may trigger execution |
AEP-1 scope
- Asset: USDC only
- Venue: Aave v3 Pool on Base (deposits earn lending yield)
- Actions:
1SUPPLY (USDC to Aave),2WITHDRAW (Aave to treasury),3TRANSFER (treasury to fixed beneficiary) - Caps: every execution <= min(2% of treasury, 100 USDC), evaluated at execution time
- Timing: 24h timelock, 48h execution window, 24h cooldown, one live proposal at a time
The TRANSFER action is the only way funds leave, and only ever to one immutable
address - so funds are never permanently locked, but exits are slow, capped, and
public. Changing any rule means deploying a new immutable version.
Live deployment (Base mainnet)
| Contract | Address |
|---|---|
| Constitution | 0xb1CE4b0007A95b1D4E6Bf57Ce258a8E8cD8d9578 |
| ProposalRegistry | 0x842259F0Ef3acc932d307Ae477E3cEa779303574 |
| Router (treasury) | 0xF6FE087817647aB39d11E8F69d71AE94A31FC97F |
All three are verified on BaseScan. Ethan's first constitutional action - a 0.20 USDC supply into Aave, submitted at the 2% cap - was executed on-chain after surviving its full 24-hour timelock (which correctly rejected early execution attempts first).
Security model
The Router is structurally incapable of anything except the three actions above: it can only call the one Aave pool and the USDC token, can only approve the pool (allowance zeroed after each use), and can only pay the one fixed beneficiary. The worst case under a fully malicious proposer is a slow, capped, publicly announced drain toward that known address - never to an attacker, never fast, never more than min(2% treasury, 100 USDC) per 24 hours, with a 24-hour veto window on every proposal.
No deployed contract is "hack proof." Ethan's claim is narrower: it is small enough to read in one sitting (~330 lines), heavily tested, statically analyzed (Slither clean of high/medium findings), and scoped so tightly that the blast radius of failure is bounded and observable. Its single external dependency is Aave v3.
Testing
Install forge-std, build, and run the suites:
forge install foundry-rs/forge-std
forge build
forge test -vv
forge test --match-path "test/fork/*" -vv
The suite covers the proposal lifecycle, caps, timelock, cooldown, veto, guardian renounce, reentrancy, and fund-conservation invariants (held across tens of thousands of randomized operations). The fork tests run the real contracts against the live Aave pool on Base.
Deploy
Fill .env from .env.example (mainnet addresses are already included), then:
forge script script/Deploy.s.sol --rpc-url https://mainnet.base.org --account <keystore> --broadcast --verify
A Base Sepolia rehearsal script (script/DeployTestnet.s.sol) deploys mocks with
short timings for testing.
Interacting
Submit a proposal to supply 10 USDC (action 1, amount in 6-decimal units):
cast send <REGISTRY> "submit(uint8,uint256)" 1 10000000 --rpc-url https://mainnet.base.org --account <keystore>
Execute it after the timelock (anyone can):
cast send <ROUTER> "execute(uint256)" <id> --rpc-url https://mainnet.base.org --account <keystore>
Proposals can be generated by anything - a human, a script, an LLM. The protocol neither knows nor cares, because proposers have no power.
Links
- Website: kaymyg.brave (IPFS)
- X: https://x.com/kay_myg
- Hugging Face: https://huggingface.co/Sahek
- BaseScan: https://basescan.org/address/0xF6FE087817647aB39d11E8F69d71AE94A31FC97F
Contact
Maintainer: kaymyg / kaymyg@proton.me
License
MIT
Not financial advice. Ethan accepts no deposits; funds sent to it cannot be recovered by the sender.
- Downloads last month
- 62