Hyperledger Cacti v3: Open-source Infrastructure for Cross-Chain Research and Financial Interoperability

Blockchain interoperability has moved from a research curiosity to a hard requirement. Regulated financial institutions, CBDC pilots, tokenization platforms, and supply-chain consortia are all converging on the same question: how do we move value, data, and trust across independently operated ledgers — safely, verifiably, and in a way that can be standardized?

Hyperledger Cacti helps achieve that. With the v3 release we are re-introducing Cacti not just as a set of connectors, but as general-purpose infrastructure for cross-chain research and financial interoperability. Most notably, we have contributed Typescript and Rust implementations of the IETF Secure Asset Transfer Protocol (SATP).

Hyperledger Cacti v1 brought the basic infrastructure "many connectors under one project.” v2 consolidated the plugin architecture, and now v3 is the release where Cacti grows up to be stable, use-case driven interoperability infrastructure.

The Secure Asset Transfer Protocol (SATP) is the strategic centerpiece of Cacti's cross-chain story. The SATP working group has continued to advance through the IETF process. Core and architecture documents are progressing through IESG evaluation, moving SATP toward RFC status as an open, standards-track protocol for gateway-to-gateway asset transfers. Our implementations are the primary open source reference implementations of SATP.

Overview of Feature Work Since v2.1.0

v3.0.0 is the largest release in Cacti's history — 18 months of sustained development, 42 contributors, and two new plugins (BUNGEE-Hermes, COPM-Corda) delivering production-grade cross-chain interoperability. The v3.0.1 patch (Aug 21) addressed minor issues in Hyperledger Fabric connectors and Weaver SDKs.

The v3.0.0 line is a feature-forward release. The most impactful new capabilities:

• SATP Hermes matures.
1. New adapter layer decouples SATP protocol logic from ledger backends, making it much easier to plug new chains into cross-chain transfers.
2. Improved gateway configuration with a cleaner, more expressive config object.
3. New get ledgers endpoint and expanded audit endpoint capabilities for observability and compliance.
4. Persisted oracle logs, giving durable, replayable evidence for oracle-mediated cross-chain flows.
5. Docker images and npm packages for every release and dev version.
• Connectivity expands.
1. Two brand-new cross-chain plugins (BUNGEE-Hermes and COPM Corda)
2. Real-time event streaming for both Fabric and Ethereum,
3. Many others; please check the official technical release.

The Cleanup Initiative

Interoperability projects accumulate surface area fast — many ledgers, many SDKs, many dependencies. A significant, deliberately large chunk of v3 is a structured cleanup initiative paying that debt down. In summary:

• Deprecated/archived packages. Retiring code that no longer has active maintainership or real-world usage, so the community's energy focuses where it matters. Ongoing discussion around further pruning and updating (e.g., older Corda/Fabric connector variants) continues in the open.
• Repository restructuring. A parallel effort is moving heavyweight tests and example packages into a dedicated cacti-demos repository so the core monorepo stays focused on shippable components.
• Documentation revamp. A coordinated documentation program is underway: architecture refresh, contributor onboarding rewrite, standardized package/connector/tooling/example docs, and MkDocs publishing.
• Security and supply-chain hygiene. CodeQL high-severity findings resolved, transitive vulnerabilities cleared, and many other security fixes.
• Modernized tooling. CI runners refreshed, affected-package computation fixed, dependency groups pinned across npm/Cargo/Go modules.
• Contributor experience. New/refreshed guidelines (CONVENTIONS.md, AI_GUIDELINES.md, PULL.md), improved pre-commit hooks, and clearer path-scoped rules for TypeScript, tests, plugins, connectors, and OpenAPI work.
  

Project Health

Cacti is stronger than ever as development velocity accelerated sharply this quarter, which culminated with the v3 release. Pull requests opened doubled from 158 to 316 (+100%), while merged PRs climbed from 65 to 97 (+49%). This marks the busiest quarter of the past year by PR volume (and one of the busiest of the project’s lifetime), closely aligned with the ongoing v3 push.

The surge in activity signals strong momentum and heightened engagement across the codebase. The issue backlog is trending in the right direction. Issues opened dropped ~46% (from 68 to 37), while closures rose ~28% (from 40 to 51). The result is a net decline in open issues, even as triage activity continues. This indicates the project is not just keeping pace but actively working down its backlog rather than accumulating unresolved work.

The contributor base is broadening at the edges while remaining concentrated at the core. With 51 active contributors and a 38% quarter-over-quarter retention rate, the project welcomed a notable cohort of highly active newcomers—including ParthSinghPS, mn-ram, malsomesh9, abhayrajjais01, futurehua, oindrilakha12-ui, Rahul-2k4, and Dev10-sys. New contributors landed a meaningful share of fixes in SATP-Hermes and the connectors, a healthy sign for long-term sustainability.

Responsiveness is stable but has clear room to improve. The ~14-day average PR merge lead time is healthy for a monorepo of this scale, and activity was recorded on roughly 85 of 90 days. However, the median issue resolution time exceeding 90 days stands out as the most obvious target for improvement next quarter. Community reach remains solid, with 414 stars, 362 forks, and 414 watchers.

Overall assessment: Healthy. LFX Insights rates the project as healthy, and the data supports that view as the trend comprises a mix of rising velocity, a shrinking backlog, and an expanding contributor base. The primary area to improve going forward is issue resolution latency, which remains the clearest weak spot amid otherwise strong indicators.

What's Next

Hyperledger Cacti v3 roadmap is still being worked on. Some of the ideas for the next cycle are:

• Keeping on track with the newest versions of SATP, which is developed toward RFC status
• Expanding our SATP implementation audit, oracle, deployment and observability features for enterprise-grade deployments.
• Deeper alignment with LFDT-wide interoperability efforts and continued academic collaborations.
• More end-to-end examples that show Cacti as production infrastructure.

We are welcoming contributions and experimentation! To contribute, please refer to https://hyperledger-cacti.github.io/cacti/contributing/how-to-contribute/

Acknowledgments

We would like to thank André Augusto, Carlos Amaro, and Michał Bajer for significant contributions to the v3 release and for being part of the team. We would like to thank the LFDT staff for their continuous support, with a special note to Ry Jones, for kindly supporting us over and over. We would also like to thank new contributors, in particular Parth Singh for helping improve the project.