Cardano Java Stack Automates Transaction Recovery With TxFlow
Cardano Client Lib is adding portable transaction plans, automatic UTxO chaining and durable recovery. Yaci Store and the pre-release Yano data node extend the same Java-first stack into indexing and local network testing.
By SongMarketCap
Updated:
Cardano Community has published the 71st Developer Office Hours, where Cardano Foundation Lead Blockchain Architect Satya Ranjan presented new tooling for Java developers building on Cardano.
Cardano Client Lib is available as the stable 0.7.2 release, while the pre-release line has reached 0.8.0-pre5. The newer components target multi-step execution, rollback handling and larger transaction workloads.
Cardano TxPlan Moves Transaction Logic Into YAML
Cardano Client Lib is a pure Java SDK supporting address derivation, payments, token minting, Plutus scripts, staking and governance operations. It does not require native libraries and can also work with JVM languages such as Kotlin and Scala.
The new TxPlan layer allows transaction instructions to be defined in portable YAML or JSON documents instead of remaining entirely inside application code. A plan can describe payments, minting and script interactions while addresses, signing keys and policy references are supplied securely at runtime.
Developers can reuse the same plan with different recipients, accounts or token parameters. The definition can also be serialized and stored, preserving the information required to rebuild a transaction after a rollback or interrupted execution.
TxPlan connects directly to Cardano Client Lib’s QuickTx builder. Existing transaction composition and provider infrastructure remain available, while repeatable business instructions move into a format that can be validated and executed independently.
TxFlow Handles Chaining, Rollbacks and Transaction Streams
TxFlow coordinates processes requiring several dependent blockchain transactions. A developer can define steps such as deposit, distribution and withdrawal, then specify which outputs from an earlier step should fund the next one.
The engine supports sequential processing, pipelined chaining and parallel batches. Pipeline mode allows a later transaction to consume an output created by a pending transaction without waiting for a separate block confirmation.
Confirmation requirements, retry policies and rollback recovery are configured inside the flow. This removes the need for each application team to build its own monitoring and resubmission layer.
TxFlowStream extends the model to continuous and high-volume workloads. Ranjan demonstrated CSV input and explained that an open stream could also receive requests through an HTTP endpoint or external adapter. Kafka is not included in the library but can feed instructions into the stream from outside.
Separate lanes allow different funding addresses to process transactions in parallel while maintaining first-in, first-out execution within each lane. Business identifiers prevent duplicate execution, while H2 and PostgreSQL storage allow processing to resume after a restart.
The demonstrations covered ADA payments, Plutus script execution, token minting and batched ADA payouts. They also included automatic UTxO chaining and recovery after a simulated interruption.
Yano Expands Cardano’s Java Testing Infrastructure
The presentation also covered Yaci Store 3.0.0-beta3, the next major beta line of the modular Cardano indexer. The release expands Blockfrost-compatible APIs and includes governance and Protocol Version 11 corrections, although complete Blockfrost compatibility has not yet been reached.
Yano extends the Java stack into node and testing infrastructure. The Cardano data node is built on Yaci and Cardano Client Lib, uses RocksDB for persistence and can run as a standalone Quarkus application or an embedded Java library.
Within Yaci DevKit, Yano can initialize and advance a local Cardano network before transferring block production to the Haskell cardano-node. A Yano-only mode is also available when developers need a lightweight environment for application and SDK testing.
Yano remains a pre-release project and is not intended for production validation. Ranjan also previewed application-specific state machines whose Merkle roots can be anchored to Cardano, but that capability remains experimental.
The three components now have distinct roles within the Java stack. Cardano Client Lib defines and executes transactions, Yaci Store indexes network data, and Yano supplies a controllable node environment for testing. Java teams can replace custom rollback services and manual local-node orchestration with reusable infrastructure before their applications reach production.