feat: setup by igorroncevic · Pull Request #2 · cowprotocol/solver-7702-delegate

igorroncevic · 2026-05-14T11:21:43Z

Description

This PR sets up the first version of Solver7702Delegate, an ERC-7702 delegation target for solver EOAs.

The contract lets approved auxiliary EOAs submit calls through the solver EOA. This gives solvers more nonce lanes for settlement submission while keeping the solver EOA as the address seen by downstream contracts.

What changed

Added Solver7702Delegate.
Added five immutable approved caller addresses set at deployment time.
Added fallback-based call forwarding.
Uses packed calldata: the first 20 bytes are the target address, and the rest is forwarded as calldata.
Forwards ETH and bubbles up target return data or revert data.
Rejects callers that are not in the approved caller list.
Added a deploy script for Solver7702Delegate.
Removed the starter Counter contract, script, and test.
Updated README, Foundry, Slither, Solhint, and editor config for this contract.

Why

A solver EOA has one nonce lane, so transactions from that EOA must land in order. With ERC-7702, auxiliary EOAs can send transactions to the solver EOA and use their own nonce lanes, while calls still execute as the solver EOA.

This keeps the existing solver identity and avoids having to allowlist several EOAs or require every solver to use a smart contract account.

Out of scope

Full test coverage. This is added in PR tests: add solver delegate coverage #3.
More detailed docs.

Testing

Run:

just build
just lint
just slither

gemini-code-assist

Code Review

This pull request replaces the example Counter contract with Solver7702Delegate, an ERC-7702 delegation target that forwards calls from authorized addresses. It also updates tool configurations for Solhint, Slither, and Foundry. Feedback identifies a compilation error in the target address extraction logic and suggests using granular Slither suppressions instead of global detector exclusions to preserve security analysis depth.

igorroncevic · 2026-05-14T11:30:49Z

+    "func-visibility": ["error", { "ignoreConstructors": true }],
+    "no-complex-fallback": "off",
+    "no-inline-assembly": "off"


All of these are necessary:

"func-visibility": ["error", { "ignoreConstructors": true }] -> constructor doesn't require explicit visibility modifier since Solidity 0.7

"no-complex-fallback": "off" -> we explicitly need our fallback to be complex

"no-inline-assembly": "off" -> we explicitly need assembly for gas efficiency

id almost say that no-inline-assembly should be added to the template repo because we almost end up using it for one reason or another.

I'd sit on that as it might not be required by other repos.

igorroncevic · 2026-05-18T09:03:21Z

+    /// @notice Address of the first approved caller
+    address private immutable APPROVED_CALLER_0;
+
+    /// @notice Address of the second approved caller
+    address private immutable APPROVED_CALLER_1;
+
+    /// @notice Address of the third approved caller
+    address private immutable APPROVED_CALLER_2;
+
+    /// @notice Address of the fourth approved caller
+    address private immutable APPROVED_CALLER_3;
+
+    /// @notice Address of the fifth approved caller
+    address private immutable APPROVED_CALLER_4;


I left these as private as to not increase the bytecode of the contract unnecessarily.

However, there would need to be some additional assumptions in services about this, because now we can no longer check if the caller is approved with isApprovedCaller (first iteration), but through some other means, e.g. simulating a transaction with a sender being e.g. Bob, and see if it would revert with Unauthorized.

I don't think this is a major issue because if the contract is deployed with CREATE2, then the correct list of callers can be validated by checking the existance of the code at the expected address.

Most ethereum libraries provide some sort of a function to predict the create2 address, such as get_create2_address_from_hash. No relying on simulation/revert on creation check needed.

To then know which addresses are authorized, the list can be read from the driver instance that created it.

kaze-cow

I just feel like the deploy script is a bit more complicated than it needs to be...

kaze-cow · 2026-05-19T05:25:13Z

+import {Solver7702Delegate} from "../src/Solver7702Delegate.sol";
+
+/// @title DeploySolver7702Delegate
+/// @author CoW Foundation


in the past we have used CoW DAO developers most frequently ex.

https://github.com/cowprotocol/euler-integration-contracts/blob/master/src/CowWrapper.sol#L16
https://github.com/cowprotocol/flash-loan-router/blob/main/src/FlashLoanRouter.sol#L13

(Other contracts have used "CoW Swap Developers" or CoW Developers, so at the very least lets not add a whole nother author to the mix)

As per cowprotocol/contracts-template#20 and legal recommendations, CoW Foundation is the name we should use going forward.

kaze-cow · 2026-05-19T06:01:02Z

-This project is meant to be used as a templated during the creation of new Github repositories (will show in the `Create a new repository > Configuration > Start with a template` selector).
-
-It will contain some useful configuration files and scripts, that can be used also with existing projects (manually copied).
+Read more about the initiative [here](https://www.notion.so/cownation/Solver7702Delegate-Design-Doc-3588da5f04ca80a1b521c436abf17724).


linking to a private notion on a public repo's readme seems a bit wierd. not sure if we want this here or not.

Good point, I feel like it should be there for any internal integrators.

anxolin

Overall looks great.

anxolin · 2026-05-19T12:28:32Z

+    /// @notice Predicts the CREATE2 address using environment variables.
+    /// @dev Reads `SALT`, approved callers, and optional `CREATE2_DEPLOYER`.
+    /// @return The predicted Solver7702Delegate address.
+    function predictAddress() external view returns (address) {


why not using this for validation during the deployment?

Validation/assumptions should now be done manually if --broadcast is not supplied, for simplicity.

kaze-cow

great!

* tests: wip * better tests * unit tests * refactor * integration tests * fork tests * check snapshots in CI * add other networks * more precise snapshot * test: address PR review comments * add rpc urls to env * fix tests * simplify tests * pr feedback * remove OZ, update tests * apply copilot recommendation to use msg.value appropriately in historical txs * update snapshots * pr feedback * add swap and bridge tx

feat: setup

58f2f5d

gemini-code-assist Bot reviewed May 14, 2026

View reviewed changes

Comment thread src/Solver7702Delegate.sol

Comment thread slither.config.json

igorroncevic marked this pull request as ready for review May 14, 2026 11:26

igorroncevic requested a review from a team as a code owner May 14, 2026 11:26

igorroncevic commented May 14, 2026

View reviewed changes

update README

e75ef0d

igorroncevic assigned igorroncevic and unassigned igorroncevic May 14, 2026

update README

7ea2f7b

igorroncevic force-pushed the feat/setup branch from cf9c6c6 to 7ea2f7b Compare May 14, 2026 13:18

igorroncevic added 2 commits May 15, 2026 14:55

Update README.md

cff45c1

added deploy script

87e139b

igorroncevic commented May 18, 2026

View reviewed changes

kaze-cow requested changes May 19, 2026

View reviewed changes

anxolin reviewed May 19, 2026

View reviewed changes

igorroncevic added 2 commits May 20, 2026 11:14

Merge branch 'main' into feat/setup

6aff2a5

pr feedback

76fd6ff

kaze-cow approved these changes May 20, 2026

View reviewed changes

igorroncevic merged commit acde4b0 into main May 21, 2026
6 checks passed

igorroncevic deleted the feat/setup branch May 21, 2026 11:34

Conversation

igorroncevic commented May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

What changed

Why

Out of scope

Testing

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kaze-cow May 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kaze-cow left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

anxolin left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

kaze-cow left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

igorroncevic commented May 14, 2026 •

edited

Loading

kaze-cow May 19, 2026 •

edited

Loading