feat(contract): API to submit foreign chain config without RPC url consensus by DSharifi · Pull Request #2784 · near/mpc

DSharifi · 2026-04-09T14:16:36Z

This PR allows nodes to submit their RPC configuration on chain, and exposes a new API to get supported chains which doesn't require all nodes to use same RPC urls. Instead a chain is considered "supported" as long as all nodes have a configuration of it, regardless of their RPC urls.

The previous API to submit "policies" will be deprecated in a follow up PR.

For all the changes that will be done see #2779

claude · 2026-04-09T14:27:12Z

Code Review

The PR adds a new ForeignChainSupport data model allowing nodes to independently register their RPC configurations, with get_supported_foreign_chains returning the intersection (chains supported by all active participants). The migration from the old policy model is well handled. Good test coverage across unit, and sandbox tests.

One concern (non-blocking):

Migration panics on non-Running state (v3_8_1_state.rs:46): The From<MpcContract> migration panics if the contract isn't in Running state. This is fine as long as upgrades are always gated on Running state, which appears to be the convention. Just flagging for awareness — if an upgrade were ever triggered during Resharing/Initializing, it would brick the migration.

No critical issues found. The logic for unanimous support checking is correct (is_superset against active participants), the backward-compat bridge in vote_foreign_chain_policy properly populates the new data model, and the TODO for cleaning stale votes after resharing is appropriately tracked.

✅ Approved

DSharifi · 2026-04-09T14:29:18Z

One concern (non-blocking):

Migration panics on non-Running state** (v3_8_1_state.rs:46): The From<MpcContract> migration panics if the contract isn't in Running state.

Migrations always happen in running state. We always use this pattern in the From implementation for previous states.

netrome

Mostly looks good to me but I find the votes_per_chain name misleading which made me confused when I initially read this.

kevindeforth · 2026-04-10T12:57:36Z

There is quite a bit of code duplication in this file.
It would be nice to introduce helper methods for registering and chain configurations and fetching supported chains, such that we reduce boilerplate and make it easier to write or extend tests.

kevindeforth · 2026-04-10T13:00:35Z

+
+    // Then: only Bitcoin is in the supported set (Starknet is not unanimous)
+    let supported: Vec<String> = contract
+        .view("get_supported_foreign_chains")


Lets replace this string with GET_SUPPORTED_FOREIGN_CHAINS from the contract interface crate.

kevindeforth · 2026-04-10T13:11:22Z

+    fn test_migration_derives_supported_foreign_chains_from_policy() {
+        let policy = make_policy(vec![
+            dtos::ForeignChain::Bitcoin,
+            dtos::ForeignChain::Ethereum,
+        ]);
+
+        let supported: dtos::SupportedForeignChains = policy
+            .chains
+            .keys()
+            .copied()
+            .collect::<BTreeSet<_>>()
+            .into();
+
+        assert!(supported.contains(&dtos::ForeignChain::Bitcoin));
+        assert!(supported.contains(&dtos::ForeignChain::Ethereum));
+        assert_eq!(supported.len(), 2);
+    }
+
+    #[test]
+    fn test_migration_derives_empty_supported_foreign_chains_from_empty_policy() {
+        let policy = dtos::ForeignChainPolicy {
+            chains: BTreeMap::new(),
+        };
+
+        let supported: dtos::SupportedForeignChains = policy
+            .chains
+            .keys()
+            .cloned()
+            .collect::<BTreeSet<_>>()
+            .into();
+
+        assert!(supported.is_empty());
+    }
+
+    #[test]
+    fn test_migration_supported_foreign_chains_preserves_all_chain_keys() {
+        let all_chains = vec![
+            dtos::ForeignChain::Solana,
+            dtos::ForeignChain::Bitcoin,
+            dtos::ForeignChain::Ethereum,
+            dtos::ForeignChain::Base,
+            dtos::ForeignChain::Bnb,
+            dtos::ForeignChain::Arbitrum,
+        ];
+        let policy = make_policy(all_chains.clone());
+
+        let supported: dtos::SupportedForeignChains = policy
+            .chains
+            .keys()
+            .cloned()
+            .collect::<BTreeSet<_>>()
+            .into();
+
+        assert_eq!(supported.len(), all_chains.len());
+        for chain in &all_chains {
+            assert!(supported.contains(chain));


I must be missing something. We are not testing actual migration logic here, but verifying properties of the output of make_policy, which is a function defined in this test module.

If we want to test the actual migration logic, we need to construct an instance of MpcContract and test the migration method above, or break the relevant logic out to a function and test that in isolation.

kevindeforth · 2026-04-10T13:22:07Z

        self.foreign_chain_policy_votes.to_dto()
    }

+    pub fn get_supported_foreign_chains(&self) -> dtos::SupportedForeignChains {


Do we have benchmarks for the gas costs of this method?
Did we consider alternative data structures for foreign_chain_configuration_by_node?

It's currently IterableMap<dtos::AccountId, dtos::ForeignChainConfiguration>, which is a nested map:
Participant <-> { ForeignChain <-> Set<RpcProvider>> }, which essentially is: AccountId <-> {Enum <-> Set<String> }

It seems to me this is optimizing for insertion, but not for lookup.
Right now, we only use it to check whether a specific foreign chain is supported by all participants, which would be more suitable for IterableMap<ForeginChain, BTreeSet<AccountId>>.

~~It's a read method. We don't use gas for reads~~

Edit: It's used in the verify foreign transaction function

kevindeforth

some comments, but nothing blocking.

netrome

Thanks for updating!

DSharifi added 4 commits April 9, 2026 16:05

mpc-contract-changes-only

4a90c1a

remove all deprecations

7675a23

undo name position

3c3e7ea

cargo fmt

606d7ad

DSharifi marked this pull request as ready for review April 9, 2026 14:25

remove deprecation from ABI

9cbdacc

netrome requested changes Apr 10, 2026

View reviewed changes

Comment thread crates/contract/src/lib.rs

Comment thread crates/contract/src/lib.rs Outdated

DSharifi added 3 commits April 10, 2026 14:20

rename votes/support to configuration

5751482

rename to foreign_chain_configuration_by_node

cd4f712

update snapshot

e82183c

DSharifi requested a review from netrome April 10, 2026 13:31

kevindeforth reviewed Apr 10, 2026

View reviewed changes

kevindeforth approved these changes Apr 10, 2026

View reviewed changes

netrome approved these changes Apr 10, 2026

View reviewed changes

DSharifi added this pull request to the merge queue Apr 13, 2026

Merged via the queue into main with commit ac040a0 Apr 13, 2026
27 checks passed

DSharifi deleted the dsharifi/add-rpc-url-to-config-on-chain-contract-only branch April 13, 2026 12:15

DSharifi mentioned this pull request Apr 15, 2026

nodes use new API for submitting which foreign chains they support #2913

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(contract): API to submit foreign chain config without RPC url consensus#2784

feat(contract): API to submit foreign chain config without RPC url consensus#2784
DSharifi merged 8 commits intomainfrom
dsharifi/add-rpc-url-to-config-on-chain-contract-only

DSharifi commented Apr 9, 2026 •

edited

Loading

Uh oh!

claude bot commented Apr 9, 2026

Uh oh!

DSharifi commented Apr 9, 2026

Uh oh!

netrome left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

kevindeforth Apr 10, 2026

Uh oh!

kevindeforth Apr 10, 2026

Uh oh!

kevindeforth Apr 10, 2026 •

edited

Loading

Uh oh!

Uh oh!

kevindeforth Apr 10, 2026

Uh oh!

DSharifi Apr 10, 2026 •

edited

Loading

Uh oh!

kevindeforth left a comment

Uh oh!

netrome left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

DSharifi commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

claude bot commented Apr 9, 2026

Code Review

Uh oh!

DSharifi commented Apr 9, 2026

Uh oh!

netrome left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

kevindeforth Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

kevindeforth Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

kevindeforth Apr 10, 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!

Uh oh!

kevindeforth Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

DSharifi Apr 10, 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!

kevindeforth left a comment

Choose a reason for hiding this comment

Uh oh!

netrome left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

DSharifi commented Apr 9, 2026 •

edited

Loading

kevindeforth Apr 10, 2026 •

edited

Loading

DSharifi Apr 10, 2026 •

edited

Loading