feat(execution): create a package metadata object at publish time #9296

miker83z · 2025-11-15T14:28:54Z

Description of change

This PR removes the previous method for verifying the authenticate functions which worked at runtime. Then, it introduces the creation of a Package Metadata immutable object during a package publishing/upgrade. Finally, it allows to use such metadata to create AuthenticatorInfoV1 instances for accounts.

Links to any relevant issues

Fixes #8861
Fixes #8862

How the change has been tested

Basic tests (linting, compilation, formatting, unit/integration tests)
Patch-specific tests (correctness, functionality coverage)
I have added tests that prove my fix is effective or that my feature works
I have checked that new and existing unit tests pass locally with my changes

vercel · 2025-11-15T14:28:59Z

The latest updates on your projects. Learn more about Vercel for GitHub.

6 Skipped Deployments

Project	Deployment	Preview	Updated (UTC)
apps-backend	Ignored	Preview	Nov 24, 2025 8:46pm
apps-ui-kit	Ignored	Preview	Nov 24, 2025 8:46pm
iota-evm-bridge	Ignored	Preview	Nov 24, 2025 8:46pm
iota-multisig-toolkit	Ignored	Preview	Nov 24, 2025 8:46pm
rebased-explorer	Ignored	Preview	Nov 24, 2025 8:46pm
wallet-dashboard	Ignored	Preview	Nov 24, 2025 8:46pm

crates/iota-protocol-config/src/lib.rs

Dkwcs · 2025-11-17T19:56:25Z

Correct me if I'm wrong, are there any possible exploits present in the runtime ?
We verify metadata during the publishing, but can we trust these data on 100% ?

iota-execution/latest/iota-verifier/src/authenticator_verifier.rs

crates/iota-framework/packages/iota-framework/sources/package_metadata.move

…d derivation

valeriyr · 2025-11-20T14:16:47Z

Cargo.lock

 version = "3.0.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f288b0a4f20f9a56b5d1da57e2227c661b7b16168e2f72365f57b63326e29b24"
+source = "git+https://github.com/iotaledger/iota-sim.git?branch=tokio-1.46.1#20adc0b089bf9072268eea0f059fb8ea2461702f"


Do we expect this change, or should it be reverted?
I see several links updated to iota-sim.

I pushed the changes after running a sim test and without re-running cargo check or similar. Fixed 9fc51cc

valeriyr · 2025-11-21T10:50:07Z

crates/iota-e2e-tests/tests/abstract_account/abstract_account/sources/abstract_account.move

 ///
 /// The `AuthenticatorInfo` will be attached to the account being built.
 public fun builder(
-    authenticator: AuthenticatorInfoV1<AbstractAccount>,


We discussed an option of creating an account with a framework function:

public fun create_account_v1<T: key>(obj: T, info: AuthenticatorInfoV1<T>) { create_account_impl(obj, info) }

So, maybe it would be better to leave the AuthenticatorInfoV1 generic parameter and not change the interface of the examples?
We don't really need to use PackageMetadataV1 here; it makes the account implementation and testing much more complex. AuthenticatorInfoV1 makes sense because we can easily limit the authenticator type that can be attached to an account.
Even if we leave the proof for a while, it will be cleaner to leave the generic parameter.

To talk about

valeriyr · 2025-11-21T10:53:14Z

crates/iota-framework/packages/iota-framework/sources/account.move

-): AuthenticatorInfoV1<Account> {
+public fun rotate_auth_info_v1(account_id: &mut UID, proof: AuthenticatorInfoV1CompatibilityProof) {


What is the reason for removing the return value?

The return was introduced because drop was not used as ability for AuthenticatorInfoV1 . So, rotate_auth_info_v1 was never modified after adding drop to the abilities.

crates/iota-framework/packages/iota-framework/sources/package_metadata.move

crates/iota-protocol-config/src/lib.rs

valeriyr · 2025-11-21T11:23:30Z

crates/iota-types/src/move_package.rs

+            };
+
+            for (fn_name, account_type) in fns_metadata {
+                value.authenticator_metadata.push(AuthenticatorMetadataV1 {


Vec can't be used here as is; we need to add validation that it doesn't contain duplicates.
There is no guarantee that it doesn't; in this struct we don't know from where this information is taken.

Moved the creation up, where it is sure that the verifier was run. Then I added a check for duplicates in the verifier 6878cb6

valeriyr · 2025-11-21T11:29:25Z

iota-execution/latest/iota-adapter/src/programmable_transactions/execution.rs

+    /// wrapper using the collected module metadata, storage ID, runtime ID,
+    /// and package version and finally freezes it. If no relevant metadata
+    /// is found, the function exits without creating any package metadata.
+    fn create_and_freeze_package_metadata_if_present(


Suggested change

fn create_and_freeze_package_metadata_if_present(

fn create_and_freeze_package_metadata_v1_if_present(

Why? The inner logic can be easily modified to support a v2 in the future, I don't see necessary the creation of create_and_freeze_package_metadata_v2_if_present in that case.

crates/iota-types/src/move_package.rs

Dkwcs · 2025-11-21T12:06:40Z

crates/iota-framework/packages/iota-framework/sources/account.move

-/// `Account` is a phantom type representing the account type which can be authenticated.
 #[allow(unused_field)]
-public struct AuthenticatorInfoV1<phantom Account: key> has copy, drop, store {
+public struct AuthenticatorInfoV1 has copy, drop, store {


What did we decide about renaming this to more suitable name?
Like AuthenticatorDescriptor or similar?

Should be done in a different PR!

crates/iota-framework/packages/iota-framework/sources/account.move

Dkwcs · 2025-11-21T12:13:00Z

crates/iota-framework/packages/iota-framework/sources/package_metadata.move

+    /// The object id of the runtime package metadata object is derived from
+    /// this value.
+    storage_id: ID,
+    /// Runtime ID of the package represented by this metadata


Is this comment useful in this context?
Could we elaborate the meaning or remove it?

Dkwcs · 2025-11-21T12:14:40Z

crates/iota-framework/packages/iota-framework/sources/package_metadata.move

+use std::type_name::TypeName;
+
+/// Key type for deriving the package metadata object address
+public struct PackageMetadataKey has copy, drop, store {}


Is it should be V1 as well?

Nope, this can be used without versioning as we won't have the need to have 2 different keys for 2 different versions in the future.

crates/iota-framework/packages/iota-framework/sources/package_metadata.move

miker83z added 8 commits November 15, 2025 10:30

refactor(natives): remove runtime authenticate function validation

d10f928

feat(execution): create a package metadata object at publish time

915f310

feat(framework): package metadata module

e5d19de

fix(move): update examples and tests

0f88cb4

fix: update snapshots

9e38b14

fix(framework): account test

c56f321

fix(verifier): check first parameter

253ce1b

fix(e2e): update aa tests

938574c

miker83z requested review from Dkwcs and valeriyr November 15, 2025 14:28

miker83z self-assigned this Nov 15, 2025

iota-ci added sc-platform Issues related to the Smart Contract Platform group. vm-language Issues related to the VM & Language Team labels Nov 15, 2025

Dkwcs reviewed Nov 17, 2025

View reviewed changes

crates/iota-protocol-config/src/lib.rs Outdated Show resolved Hide resolved

Dkwcs reviewed Nov 17, 2025

View reviewed changes

iota-execution/latest/iota-verifier/src/authenticator_verifier.rs Outdated Show resolved Hide resolved

Dkwcs reviewed Nov 19, 2025

View reviewed changes

crates/iota-framework/packages/iota-framework/sources/package_metadata.move Show resolved Hide resolved

miker83z added 4 commits November 19, 2025 14:50

fix(protocol): remove check_auth_info_v1_cost_base

1bfa278

feat(verifier): check authenticator has object as first param

5e2956f

refactor(framework): package metadata

0c5cd21

feat(framework): add derived object and update the package metadata i…

aa05ddb

…d derivation

miker83z marked this pull request as ready for review November 20, 2025 11:53

miker83z requested review from a team as code owners November 20, 2025 11:53

miker83z removed the request for review from a team November 20, 2025 11:54

miker83z removed request for a team November 20, 2025 11:54

This was linked to issues Nov 20, 2025

[AA] Create PackageMetadata type #8861

Open

[AA] Adapt the iota-execution publish and upgrade flows to create a PackageMetadata immutable object #8862

Open

miker83z mentioned this pull request Nov 20, 2025

feat: investigate a solution where immutable objects are created when a package is published #8736

Closed

valeriyr requested changes Nov 21, 2025

View reviewed changes

Dkwcs reviewed Nov 21, 2025

View reviewed changes

crates/iota-framework/packages/iota-framework/sources/account.move Outdated Show resolved Hide resolved

Dkwcs reviewed Nov 21, 2025

View reviewed changes

crates/iota-framework/packages/iota-framework/sources/package_metadata.move Outdated Show resolved Hide resolved

miker83z added 4 commits November 24, 2025 20:00

fix cargo lock

9fc51cc

refactor(framework): package metadata getters and borrowers

458c47a

refactor(protocol): make package_metadata_version u64

6ec1eb0

various fixes

6878cb6

		): AuthenticatorInfoV1<Account> {
		public fun rotate_auth_info_v1(account_id: &mut UID, proof: AuthenticatorInfoV1CompatibilityProof) {

	fn create_and_freeze_package_metadata_if_present(
	fn create_and_freeze_package_metadata_v1_if_present(

feat(execution): create a package metadata object at publish time #9296

Are you sure you want to change the base?

feat(execution): create a package metadata object at publish time #9296

Conversation

miker83z commented Nov 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description of change

Links to any relevant issues

How the change has been tested

Uh oh!

vercel bot commented Nov 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Dkwcs commented Nov 17, 2025

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!

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!

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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Dkwcs Nov 21, 2025 • 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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

miker83z commented Nov 15, 2025 •

edited

Loading

vercel bot commented Nov 15, 2025 •

edited

Loading

Dkwcs Nov 21, 2025 •

edited

Loading