feat(stm): Add key registration support for pre-aggregation primitives #2932

damrobi · 2026-01-19T16:25:44Z

Content

This PR includes changes to the stm library, more specifically to the merkle tree, protocol and proof system modules.

Pre-submit checklist

Branch
- Tests are provided (if possible)
- Crates versions are updated (if relevant)
- CHANGELOG file is updated (if relevant)
- Commit sequence broadly makes sense
- Key commits have useful messages
PR
- All check jobs of the CI have succeeded
- Self-reviewed the diff
- Useful pull request description
- Reviewer requested
Documentation
- Update README file (if relevant)
- Update documentation website (if relevant)
- No new TODOs introduced

Comments

Issue(s)

Relates to #2792

fix register test

…sing it in AggregateVerificationKey

…by Snark variant

…nark variant

… variant

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+    merkle_tree_commitment: MerkleTreeBatchCommitment<D::SnarkHash, MerkleTreeSnarkLeaf>,
+}
+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {


mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+    merkle_tree_commitment: MerkleTreeBatchCommitment<D::SnarkHash, MerkleTreeSnarkLeaf>,
+}
+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {


mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

github-actions · 2026-01-19T16:36:40Z

Test Results

4 files ±0 169 suites ±0 23m 35s ⏱️ + 1m 26s
2 322 tests ±0 2 322 ✅ ±0 0 💤 ±0 0 ❌ ±0
7 311 runs ±0 7 311 ✅ ±0 0 💤 ±0 0 ❌ ±0

Results for commit b5d320a. ± Comparison against base commit 1bd88b8.

♻️ This comment has been updated with latest results.

…move batch and changed type of EligibilityValue

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {
+    /// Get the Merkle tree batch commitment.
+    pub(crate) fn get_merkle_tree_commitment(


mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {
+    /// Get the Merkle tree batch commitment.
+    pub(crate) fn get_merkle_tree_commitment(


Copilot

Pull request overview

This PR adds key registration support for pre-aggregation primitives by introducing a Snark proof system alongside the existing Concatenation proof system. The changes rename the "Future" variant to "Snark" throughout the codebase and implement the foundational structures needed for Snark-based aggregation.

Changes:

Renamed "Future" to "Snark" across all proof system enums and types for better clarity
Added Hash trait implementation for SchnorrVerificationKey to support its use in merkle trees
Introduced new Snark-specific types: AggregateVerificationKeyForSnark, MerkleTreeSnarkLeaf, VerificationKeyForSnark, and EligibilityValue

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 9 comments.

Show a summary per file

File	Description
`signature_scheme/unique_schnorr_signature/verification_key.rs`	Adds Hash trait implementation for SchnorrVerificationKey
`protocol/mod.rs`	Introduces VerificationKeyForSnark type alias
`protocol/aggregate_signature/signature.rs`	Renames Future variant to Snark in AggregateSignatureType and AggregateSignature enums
`protocol/aggregate_signature/clerk.rs`	Updates clerk to reference Snark instead of Future
`protocol/aggregate_signature/aggregate_key.rs`	Implements Snark variant for AggregateVerificationKey with serialization support
`proof_system/mod.rs`	Exports new halo2_snark module and its types
`proof_system/halo2_snark/mod.rs`	Creates new module structure for Snark proof system
`proof_system/halo2_snark/aggregate_key.rs`	Implements AggregateVerificationKeyForSnark with merkle tree commitment
`membership_commitment/merkle_tree/leaf.rs`	Adds MerkleTreeSnarkLeaf for Snark-based merkle trees
`lib.rs`	Exports new Snark types and defines EligibilityValue

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-21T14:16:10Z

mithril-stm/src/membership_commitment/merkle_tree/leaf.rs

+/// The values that are committed in the Merkle Tree for `ConcatenationProof`.
+/// Namely, a verified `BlsVerificationKey` and its corresponding stake.


The documentation comment incorrectly refers to both ConcatenationProof and BlsVerificationKey. This struct is for the Snark proof system (not Concatenation) and uses SchnorrVerificationKey (not BLS). The comment should be updated to accurately describe that this is used for the Snark proof system with Schnorr verification keys and eligibility values.

Suggested change

/// The values that are committed in the Merkle Tree for `ConcatenationProof`.

/// Namely, a verified `BlsVerificationKey` and its corresponding stake.

/// The values that are committed in the Merkle Tree for the Snark proof system.

/// Namely, a verified `SchnorrVerificationKey` and its corresponding eligibility value.

Copilot · 2026-01-21T14:16:11Z

mithril-stm/src/membership_commitment/merkle_tree/leaf.rs

+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, Serialize, Deserialize, Hash)]
+pub struct MerkleTreeSnarkLeaf(pub VerificationKeyForSnark, pub EligibilityValue);


The MerkleTreeSnarkLeaf struct lacks PartialOrd and Ord implementations that are present in MerkleTreeConcatenationLeaf. Given that both structures serve similar purposes (merkle tree leaves with verification keys and values), and that ordering is important for batch opening efficiency as explained in MerkleTreeConcatenationLeaf's documentation, MerkleTreeSnarkLeaf should also implement these traits.

Copilot · 2026-01-21T14:16:11Z

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+}
+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {
+    /// Get the Merkle tree batch commitment.


The comment states "Get the Merkle tree batch commitment" but the method name is get_merkle_tree_commitment and returns a MerkleTreeCommitment, not a batch commitment. The documentation should be corrected to accurately describe what is being returned.

Suggested change

/// Get the Merkle tree batch commitment.

/// Get the Merkle tree commitment.

Copilot · 2026-01-21T14:16:11Z

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+
+        bytes
+    }
+


The from_bytes method is missing documentation. Public methods should have doc comments explaining their purpose, parameters, return values, and potential errors.

Suggested change

/// Reconstruct an aggregate verification key for Snark from its byte representation.

///

/// # Parameters

///

/// - `bytes`: A byte slice containing the serialized form of this key, as produced by

/// [`AggregateVerificationKeyForSnark::to_bytes`].

///

/// # Returns

///

/// Returns an [`AggregateVerificationKeyForSnark`] instance on success.

///

/// # Errors

///

/// Returns an error if `bytes` do not encode a valid underlying

/// [`MerkleTreeCommitment`] and deserialization fails.

Copilot · 2026-01-21T14:16:12Z

mithril-stm/src/protocol/aggregate_signature/aggregate_key.rs

            }
            #[cfg(feature = "future_snark")]
-            AggregateVerificationKey::Future => vec![],
+            AggregateVerificationKey::Snark(_) => vec![],


The to_bytes method for AggregateVerificationKey::Snark returns an empty vector, which is inconsistent with the corresponding from_bytes implementation that expects actual data. This asymmetry will cause deserialization to fail or produce incorrect results. The Snark variant should serialize its underlying AggregateVerificationKeyForSnark data.

Suggested change

AggregateVerificationKey::Snark(_) => vec![],

AggregateVerificationKey::Snark(snark_aggregate_verification_key) => {

snark_aggregate_verification_key.to_bytes()

}

Copilot · 2026-01-21T14:16:12Z

mithril-stm/src/membership_commitment/merkle_tree/leaf.rs

+#[cfg(feature = "future_snark")]
+/// The values that are committed in the Merkle Tree for `ConcatenationProof`.
+/// Namely, a verified `BlsVerificationKey` and its corresponding stake.
+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, Serialize, Deserialize, Hash)]
+pub struct MerkleTreeSnarkLeaf(pub VerificationKeyForSnark, pub EligibilityValue);
+
+#[cfg(feature = "future_snark")]
+impl MerkleTreeLeaf for MerkleTreeSnarkLeaf {
+    fn as_bytes_for_merkle_tree(&self) -> Vec<u8> {
+        self.to_bytes()
+    }
+}
+
+#[cfg(feature = "future_snark")]
+impl MerkleTreeSnarkLeaf {
+    fn to_bytes(self) -> Vec<u8> {
+        let mut result = [0u8; 64];
+        result[..32].copy_from_slice(&self.0.to_bytes());
+        result[32..].copy_from_slice(&self.1.to_bytes_le());
+        result.to_vec()
+    }
+
+    // TODO: remove this allow dead_code directive when function is called or future_snark is activated
+    #[allow(dead_code)]
+    pub(crate) fn from_bytes(bytes: &[u8]) -> StmResult<Self> {
+        if bytes.len() < 32 {
+            return Err(anyhow!(MerkleTreeError::SerializationError)).with_context(
+                || "Not enough bytes provided to construct a Merkle tree leaf for Snark.",
+            );
+        }
+        let schnorr_verification_key = VerificationKeyForSnark::from_bytes(&bytes[..32])
+            .map_err(|_| MerkleTreeError::SerializationError)?;
+        let mut u64_bytes = [0u8; 32];
+        u64_bytes.copy_from_slice(&bytes[32..]);
+        let eligibility_value = match EligibilityValue::from_bytes_le(&u64_bytes).into_option() {
+            Some(base_field_element) => Ok(base_field_element),
+            None => Err(anyhow!(MerkleTreeError::SerializationError)),
+        }?;
+        Ok(MerkleTreeSnarkLeaf(
+            schnorr_verification_key,
+            eligibility_value,
+        ))
+    }
+}


The MerkleTreeSnarkLeaf struct and its methods (to_bytes, from_bytes) lack test coverage. Given that similar functionality for MerkleTreeConcatenationLeaf is tested in the merkle tree test suite, this new leaf type should also include tests to verify serialization/deserialization behavior and the Hash trait implementation.

Copilot · 2026-01-21T14:16:12Z

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct AggregateVerificationKeyForSnark<D: MembershipDigest> {
+    merkle_tree_commitment: MerkleTreeCommitment<D::SnarkHash, MerkleTreeSnarkLeaf>,
+}
+
+impl<D: MembershipDigest> AggregateVerificationKeyForSnark<D> {
+    /// Get the Merkle tree batch commitment.
+    pub(crate) fn get_merkle_tree_commitment(
+        &self,
+    ) -> MerkleTreeCommitment<D::SnarkHash, MerkleTreeSnarkLeaf> {
+        self.merkle_tree_commitment.clone()
+    }
+
+    /// Convert the aggregate verification key for Snark to bytes.
+    pub fn to_bytes(&self) -> Vec<u8> {
+        let mut bytes = Vec::new();
+        bytes.extend(self.merkle_tree_commitment.to_bytes());
+
+        bytes
+    }
+
+    pub fn from_bytes(bytes: &[u8]) -> StmResult<Self> {
+        let size = bytes.len();
+        let merkle_tree_commitment = MerkleTreeCommitment::from_bytes(&bytes[0..size])?;
+        Ok(Self {
+            merkle_tree_commitment,
+        })
+    }
+}


The new AggregateVerificationKeyForSnark struct and its methods (to_bytes, from_bytes, get_merkle_tree_commitment) lack test coverage. Given that similar structs like AggregateVerificationKeyForConcatenation have comprehensive tests, this new functionality should also include tests to verify serialization/deserialization and other behaviors.

Copilot · 2026-01-21T14:16:13Z

mithril-stm/src/membership_commitment/merkle_tree/leaf.rs

+        if bytes.len() < 32 {
+            return Err(anyhow!(MerkleTreeError::SerializationError)).with_context(
+                || "Not enough bytes provided to construct a Merkle tree leaf for Snark.",
+            );


The from_bytes method checks if bytes.len() < 32 but then tries to copy 32 bytes from bytes[32..] without verifying that there are at least 64 bytes available. This could cause a panic if the slice is between 32 and 64 bytes. The validation should check for bytes.len() < 64 instead.

Copilot · 2026-01-21T14:16:13Z

mithril-stm/src/proof_system/halo2_snark/aggregate_key.rs

+        let size = bytes.len();
+        let merkle_tree_commitment = MerkleTreeCommitment::from_bytes(&bytes[0..size])?;


The slice operation &bytes[0..size] is redundant since it's equivalent to just bytes. The entire byte slice is being passed to from_bytes, making the explicit slicing unnecessary.

Suggested change

let size = bytes.len();

let merkle_tree_commitment = MerkleTreeCommitment::from_bytes(&bytes[0..size])?;

let merkle_tree_commitment = MerkleTreeCommitment::from_bytes(bytes)?;

curiecrypt and others added 12 commits January 19, 2026 17:34

new sigregparty keyreg and participant

553df6c

aggregate verification_key

39ed457

fix register test

dedup test revised and ord for registration entry is updated

bec8e4e

try to/from bytes for avk

4afe35c

fix(stm): 'AggregateVerificationKey' bytes codecs

7e47f29

rev suggestions part 1

1bd88b8

feat(stm): Added the MerkleTreeSnarkLeaf variant of merkle tree leaves

6ba8f24

Added missing future_snark feature flags for new MerkleTreeSnarkLeaf

a3bab0b

feat(stm): Started adding the halo2_snark module of proo_system and u…

5fb3365

…sing it in AggregateVerificationKey

refactor(stm): replaced Future variant of AggregationVerificationKey …

a34b2c3

…by Snark variant

refactor(stm): replaced Future variant of AggregateSignatureType by S…

0c9be21

…nark variant

refactor(stm): replaced Future variant of AggregateSignature by Snark…

39d33fd

… variant

github-advanced-security bot found potential problems Jan 19, 2026

View reviewed changes

jpraynaud assigned damrobi Jan 19, 2026

fix(stm): modify commitment of AggregateVerificationKeyForSnark to re…

b5d320a

…move batch and changed type of EligibilityValue

github-advanced-security bot found potential problems Jan 20, 2026

View reviewed changes

curiecrypt requested a review from Copilot January 21, 2026 14:11

Copilot started reviewing on behalf of curiecrypt January 21, 2026 14:11 View session

Copilot AI reviewed Jan 21, 2026

View reviewed changes

curiecrypt force-pushed the curiecrypt/msnark/key-registration branch 3 times, most recently from 037d8b7 to b097651 Compare January 23, 2026 10:00

Base automatically changed from curiecrypt/msnark/key-registration to main January 23, 2026 11:33

		/// The values that are committed in the Merkle Tree for `ConcatenationProof`.
		/// Namely, a verified `BlsVerificationKey` and its corresponding stake.

		#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, Serialize, Deserialize, Hash)]
		pub struct MerkleTreeSnarkLeaf(pub VerificationKeyForSnark, pub EligibilityValue);

	/// Get the Merkle tree batch commitment.
	/// Get the Merkle tree commitment.

+    /// Reconstruct an aggregate verification key for Snark from its byte representation.
+    ///
+    /// # Parameters
+    ///
+    /// - `bytes`: A byte slice containing the serialized form of this key, as produced by
+    ///   [`AggregateVerificationKeyForSnark::to_bytes`].
+    ///
+    /// # Returns
+    ///
+    /// Returns an [`AggregateVerificationKeyForSnark`] instance on success.
+    ///
+    /// # Errors
+    ///
+    /// Returns an error if `bytes` do not encode a valid underlying
+    /// [`MerkleTreeCommitment`] and deserialization fails.

-            AggregateVerificationKey::Snark(_) => vec![],
+            AggregateVerificationKey::Snark(snark_aggregate_verification_key) => {
+                snark_aggregate_verification_key.to_bytes()
+            }

		let size = bytes.len();
		let merkle_tree_commitment = MerkleTreeCommitment::from_bytes(&bytes[0..size])?;

feat(stm): Add key registration support for pre-aggregation primitives #2932

Are you sure you want to change the base?

feat(stm): Add key registration support for pre-aggregation primitives #2932

Conversation

damrobi commented Jan 19, 2026

Content

Pre-submit checklist

Comments

Issue(s)

Uh oh!

Check warning

Check warning

Uh oh!

Uh oh!

github-actions bot commented Jan 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Test Results

Uh oh!

Check warning

Check warning

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 21, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

github-actions bot commented Jan 19, 2026 •

edited

Loading