CORE TECHNOLOGY

This week, the consensus team split the Leios prototype's database into two files, one for data that can still be rolled back and one for settled data, with certified Endorser Blocks moving from the first to the second in the background; operators can put each file on its own disk. Garbage collection now runs in two paced steps, marking rows and then deleting them in batches, instead of blocking the node. db-synthesizer builds an actual Leios chain now, filling blocks, voting on Endorser Blocks, and storing the ones it forges, and db-truncater can truncate that same Leios database and drop orphaned transactions when cutting a chain back to a given slot.

Block forging also moved out of the node kernel into its own module with named steps, same behavior but easier to instrument and reuse. Separately, the team dropped node-to-client protocol support below v23, which shipped with node 10.7.0 in March 2026, removing legacy code that no checked client still needs.

SMART CONTRACTS

This week, the Plutus team closed out several September milestones. A new with-crypto cabal flag lets Plinth and Plutus build without the system crypto C libraries, and the uplc certifier got further usability work. Other merged changes covered Data.List encoding for single-constructor ledger types, more V4 helper functions and test cases, efficiency and consistency fixes to Data.AssocMap, deserialiser optimizations, and checkpoints in the UPLC inliner.

Still in progress: cost models for keepPolicies and dropPolicies (CIP-0168), a CIP to drop the scope check during script deserialization, shape-safety for Data.AssocMap's FromData, exact arity checks in generated FromData, and continued work on built-in pattern matching (CIP-0194) and casing on built-in types.

Check out the Plutus Core team’s update for further details.

The Cardano High Assurance team worked on finalizing the release of the property-based testing extension, with its latest features, the Transaction Graph view and the coverage report, now reviewed and merged. A new tutorial for the testing tool is also available for newcomers (and their trusty LLM friends).

Over the past few weeks, the team also developed a new set of static analysis rules for plu-stan. This work is currently under review, with a new release expected next week.

On the automated formal verification side, the team has invested significant effort in a complete rework of the higher-order function formalization to boost Blaster's performance. They also fixed a few bugs that mainly affected power users working on very specific kinds of theorems. In addition, the team is improving the Z3 and CVC5 backend solvers, with several patches already merged. These patches directly benefit formal methods engineers around the world, while also delivering major improvements to the automated proof of Cardano smart contracts.

SCALING

This week, the Hydra team merged five PRs, one opened (and merged same day), 37 commits across two repos, no releases. The compiled head validator script was regenerated (a hash bump with no cost or behavioral change), and last week's fanout guard and adversarial model testing PRs were both merged.

The main fix addressed a fund-lockout: a user's deposit through the Mesh SDK went unobserved and sat stuck at the deposit script for three hours after coverFee topped up an under-minimum output but left its datum at the original value, so the node's value-versus-datum check rejected it. The fix validates the deposit before submission, fails fast with a new error naming the shortfall, and adds a way to decode already-stuck deposits for manual recovery.

The rest of the week went into test infrastructure: nix flake check now runs the Haskell test suites directly, and 13 slow end-to-end tests were replaced with 30 unit tests, cutting about 114 seconds of test time down to roughly 1.3 seconds. Writing the replacements exposed several things the old tests had only covered by accident, most notably that all 39 cases of which internal events reach clients, and in what form, were untested.

Read this Hydra update for more.

WALLETS

This week, the Lace team released Lace 2.3, making things feel smoother, clearer, and more reliable across the wallet. DApp connections now survive the wallet's background process restarting mid-request, with already-confirmed transactions recognized instead of reported as failures, and connected DApps correctly notified when the active account changes. The staking pool browser adds estimated annual rates and a combined recommendation score for ordering, and the DRep browser now orders by quality signals rather than stake weight, hiding policy-excluded DReps from the main list while keeping them searchable by ID.

Hardware wallet support improved across Ledger, Trezor, Keystone, and SeedSigner, including separate on-device certificates for stake registration and delegation and better handling of the extra signatures air-gapped script-based wallets need. Balance and sync displays were cleaned up so accounts show progress instead of appearing empty, and Midnight accounts gained a manual sync reset along with fixes to DUST fee calculation and a background state-write bug.

Read this blog post for more on the latest Lace release.

RESEARCH

This week, the Research team hosted an interactive technical workshop on Cardano governance. If you missed this workshop, you can watch the recording. Stay tuned for more technical workshops from IO Research.


Weekly development report as of 2026-09-25

CORE TECHNOLOGY

This week’s update from the ledger team covers four weeks of work. The headline item: Dijkstra's batch validation now runs Plutus scripts for sub-transactions and counts their execution budgets, closing a gap where sub-transaction validators and minting policies went unchecked and their costs escaped transaction and block limits and script fees. The team also reworked Withdrawals handling in the ENTITIES rule to implement the full specification, moving batch withdrawal checks out of UTXO and into ENTITIES along with legacy-mode value conservation and new SUBUTXO/SUBUTXOW predicate failures.

On Plutus V4, the team implemented the TxInfo translation for protocol version 4, redid the PlutusPurpose translation, and fixed the Dijkstra cost model injection, bumping plutus-ledger-api to at least 1.68. Dijkstra era work also added a POOL rule, an EMax /= 0 check, and enforced account balance intervals, while Leios protocol parameters landed in the Dijkstra era with a size-constrained leios\_certificate signer set.

The rest was serialisation, interface, and testing cleanup: more resilient decoders and encoders, an era parameter on StakePoolParams, ToJSON/FromJSON instances for transaction types across all eras, new conformance tests for Dijkstra's EPOCH rule and account balance intervals, and a crypton version bump alongside CI hardening and node 11.1.1 backports.

Read this update for more details. 

SCALING

This week was one of consolidation for the Hydra team after the 2.4.0/2.4.1 release: four PRs merged, three opened, across two repos. The etcd network component was bumped (etcd 3.5.25 to 3.6.6, grapesy and http2 client libs updated too); the long-standing putMessage stall under heavy load now surfaces as a reconnectable error instead of hanging forever, though the underlying issue is still reproducible upstream. The actual fix, dedicated recycled broadcast connections, was re-verified with large soak tests. A nightly Blockfrost test flake was also fixed.

The main bug fix closed a fanout wedge: a partial fanout could empty a head while its on-chain state still read Closed, leaving it stuck unable to finalize or revert. The fix centralizes the fanout decision so every caller hits the same guard. New adversarial model testing (spam, rollbacks, latency, restarts) caught and fixed a similar wedge plus several mock-chain bugs.

Separately, Performance (M2) is now marked done, and three other milestones (metrics, the hydra-recipes repo, and cardano-node compatibility) moved earlier on the roadmap.

Here’s this week’s Hydra update in greater detail.

RESEARCH

This week, the Research team is attending MARBLE 2026 to present the Strategic Validator Liquidation in Proof-of-Stake Tokenomics research paper and SCN 2026 to present the Improved Approximation Guarantees for Groupwise Maximin Share Fairness research paper.

They are also preparing to host the next in the series of technical workshops under Cardano Vision 26 on Cardano governance. Registration is now open.


Weekly development report as of 2026-09-18

CORE TECHNOLOGY

This week, the consensus team made several improvements to the Leios prototype. They rewrote the download decision logic so nodes bound both the time and memory spent on pending downloads, prioritize the newest endorser blocks, and avoid saturating the same peer with concurrent requests. They also added a transaction cache so nodes don't re-download transactions they have already seen. The team fixed a second cause of unwanted disconnects between honest nodes – syncing nodes now pause requests for Leios messages until they catch up – and capped the time the mempool spends reading a ledger state from disk, preventing block production delays.

On the storage side, ledger state garbage collection now runs in its own thread, preventing snapshot delays from blocking memory cleanup.

Finally, the team released ouroboros-consensus 4.2.0.0 and followed up with 4.2.1.0, which fixes a silent failure in the GetGenesisConfig query that had caused cardano-wallet to stop following the chain against node v.11.0.1.

See this update for more details.

SMART CONTRACTS

This week, the Plutus team merged casing on Data: a case on a Data.Constr value now dispatches directly on the constructor tag, which, combined with the recently added dropList, significantly reduces the cost of consuming Data-encoded values. This is a Plutus V4 feature, and work is also underway on specifying casing on built-in types in the Plutus Core specification.

Work on the Value built-ins (CIP-0168) continues: the team added the keepPolicies and dropPolicies built-ins and merged the cost model for policies, with cost models for keepPolicies and dropPolicies in progress. As with other new built-ins, these are gated behind a future protocol version and cannot be used on-chain until the Dijkstra era.

On the Plutus V4 ScriptContext front, the team is implementing a different encoding for product types using List instead of Constr, making them cheaper to decode and improving the usability of the certifier in the uplc and plc tools.

The Cardano High Assurance team released the first stable release of Plu-stan, the static analyzer for Plutus, with additional rules to cover more common pitfalls developers face when first developing their DApps. It comes with an updated version of the Visual Studio Code extension.

Work on the Property-based testing tool is still ongoing, with several critical features being finalized for the Visual Studio Code extension: a test tree view to quickly identify test-run results, and an intuitive transaction graph view to review failed tests at the transaction level, as a user would draw them. The extension is available for very early testing to interested users.

Finally, in the formal verification tool, the team fully axiomatized BLS functions and hashing functions to enable high-level proofs that use the expected security properties of those primitives. They are also continuously testing and pushing optimizations on lean-blaster to scale the tool to larger and larger scripts.

DEVELOPER EXPERIENCE

This week, the developer experience team delivered milestone DX.02 (the cardano-init setup CLI specifications and proof of concept) and published the first open beta release, receiving strong community support.

Work continued to harden and expand the tool selection, including ongoing integration of Dingo as a devnet provider. The contracts-library reached a significant milestone by completing two reusable protocol implementations, each delivered with on-chain and off-chain code, specifications, and examples: a settings protocol for managing any protocol's configuration, and a governance and DAO protocol for fully on-chain protocol governance.

On the Cardano Foundation's Developer Portal, nine of twelve intermediate onboarding lectures are now complete.

Dive deeper for more details.

SCALING

This week, the Hydra team released 2.4.0 and immediately followed with 2.4.1, a security release. Anyone running 2.3.0 or 2.4.0 should upgrade to 2.4.1.

2.4.0 introduced security fixes to incremental commits: a snapshot now identifies exactly which deposit it approves, and the on-chain validators enforce that an increment can only claim the deposit the parties actually signed. As this changed the Hydra scripts and snapshot signature payload, the upgrade was breaking – open heads must be closed and fanned out before upgrading, and the persisted hydra.db migrated to CBOR on first start with no downgrade path.

2.4.1 superseded 2.4.0 the day after a vulnerability was identified: when confirming a snapshot, hydra-node re-applied transactions through an optimized path that skipped signature verification and Plutus script evaluation. An unvalidated transaction could reach that path, allowing a malicious head participant to get a transaction with an invalid signature or a failing script into a confirmed snapshot and steal funds. 2.4.1 removes the optimization and fully validates all transactions when processing a snapshot request. It is a drop-in replacement for 2.4.0 with no changes to scripts, snapshot signatures, or database format.

The release also bundles this quarter's accumulated work, including the CBOR database and binary client API, Blockfrost performance improvements, Linux/arm64 Docker images, the Typst/Agda specification, and several deposit-handling and validity-window fixes.

See this update for more details.

RESEARCH

This week, the research team are preparing to present the following papers at two world-renowned conferences:

MARBLE 2026

Strategic Validator Liquidation in Proof-of-Stake Tokenomics: Lingxiao Zhao, Madalina Sas, and Carmine Ventre.

SCN 2026

Adaptively Secure Fast Settlement with Dynamic Participation and Self-Healing: Christian Badertscher, Sandro Coretti-Drayton, Peter Gaži, Aggelos Kiayias, and Alexander Russell. 

The team also hosted the data availability technical (DA) workshop, which covered on-chain DA requirements for Cardano’s growing layer 2 ecosystem. If you missed this workshop, you can watch the recording. 

Stay tuned for more technical workshops coming soon!


Weekly development report as of 2026-09-11

CORE TECHNOLOGY

Last week, the performance and tracing team benchmarked the Cardano node v.11.1.0 and analyzed both release and pre-release performance. The benchmarks showed reduced process CPU usage, alongside an increase in resident set size, which is being addressed in the v.11.1.1 patch release.

The team also enhanced beacon benchmarking with metrics that separate ledger-tick and disk-access times for UTXO lookups, and replaced its Cairo-based plotting backend with gnuplot to improve portability and reduce dependencies. For Leios, they delivered on-disk LedgerDB transaction validation benchmarks, including disk I/O, memory, and block I/O measurements, and continued work on a self-contained benchmark package that can run on SPO hardware without Nix or network access. In tracing, implementation of the cardano-tracer alert manager is now underway, following the earlier design work. 

SCALING

This week, the Hydra team made significant progress on the specification stack, moving the specification from LaTeX to Typst and formalizing it in Agda. They also added differential testing between the Hydra node and the validator, along with a machine-checked reference, strengthening confidence in the implementation and specification. The stack is part of the maintenance and developer experience work and is now integrated.

Additionally, the team improved reliability and performance across Hydra, including Blockfrost query caching, script integrity fixes, stricter UTXO value handling, and improvements to benchmark consistency. They also reduced smoke test execution time and continued progress on partial fanout, including more efficient chunk-size searches and broader property testing.

Finally, the team improved operational support by publishing native Linux ARM64 Hydra node images alongside AMD64 images and enhancing tracing around the KZG trusted setup. They also redeployed hydra-explorer to EC2 and made further improvements to documentation and testing.

For a detailed breakdown, see this technical report.

RESEARCH

This week, the research team held the second in a series of technical workshops under Cardano Vision 26, which covered data availability requirements, use cases, and specifications. It also included preparing a Cardano Problem Statement (CPS), roadmap discussions, and implementation handover. This workshop was hosted by Fergie Miller, director of research partnerships, and Giorgos Panagiotakos, research fellow at Input Output Group.


Weekly development report as of 2026-09-04

CORE TECHNOLOGY

This week, the consensus team resolved an issue with the attack in the Leios prototype targeting how nodes announce endorser blocks, which had been causing honest nodes to disconnect from each other (#2179). They also extended db-analyser to read Leios chains, enabling the tool to replay a chain and report the cost of applying each block, including the transactions of each certified endorser block, their total size, and the time to read them (#2181).

On the maintenance side, the team forward-ported the new snapshot interval and mempool allocation fix from the 4.x release line to the main development branch (#2209, #2215).

WALLETS AND SERVICES

The Lace team added support for the Keystone hardware wallet, available in both the Lace browser extension and the Lace mobile app. Keystone is an air-gapped device that uses QR codes for signing rather than USB, Bluetooth, or NFC connections, removing the attack surface associated with live data connections. Seed phrases are generated using true random number generators across three separate secure element chips, with an option for users to generate their own entropy manually using dice rolls.

Lace now supports four hardware wallets – Ledger, Trezor, SeedSigner, and Keystone – each taking a different approach to security and connectivity. Read more in the full blog post.

SMART CONTRACTS

The Plutus team published release 1.68.0.0, which includes the Plutus V4 ledger API types. Note that Plutus V4 is still under development, and these types may change before the Dijkstra era.

Getting started with Plinth is now easier: the new install script in plinth-template sets up a Plinth project with a single command and handles installing the required tooling. See the updated documentation for details.

Work on Plutus V4 language features continues. The team implemented casing on Data: a case on a Data.Constr value can now dispatch directly on the constructor tag, which, combined with the recently added dropList, yields a performance improvement. The team also added the assetCount built-in (CIP-0168) along with its cost model, merged the cost model for multiIndexArray (CIP-0156), and continued formalizing built-ins in the Agda metatheory.

Feedback on CPS-0034 – Extending Plutus Core conformance testing is welcome.

DEVELOPER EXPERIENCE

This week, the developer experience team delivered the first tagged releases of cardano-init – v0.2.0 and v0.2.1, marking the tool's transition from prototype to released product. The web builder surface was dropped, making the CLI the single, focused interface. Two new commands – cardano-init add and remove – now allow developers to edit a project's composition after generation, swapping or adding tools without starting from scratch. Network selection was also simplified: every project now generates for preview by default, with a one-line environment variable edit to switch networks at runtime.

The first external contributions landed this week, with Chris Gianelloni from Blink Labs opening several pull requests including a new Dingo devnet provider and a cluster of security and reliability fixes.

The contracts-library saw 34 commits, with the configuration parameter management pull request still in progress. The Cardano Foundation's Developer Portal onboarding track advanced with 10 further commits. The ecosystem map was updated to include team contacts and maintainer handles for easier community outreach.

As always, check out the technical report for more details. 

SCALING

The Hydra team delivered two significant performance improvements as part of the CBOR milestone. Persisted event payloads now use CBOR instead of JSON, reducing a real three-node benchmark database by 2.6× (27.4 MB to 10.5 MB), with automatic migration included. The client API also gains opt-in binary CBOR encoding, reducing wire sizes by 2.2–3.1× and speeding up encode/decode by 3–21×; a 1,000-UTXO snapshot decode drops from 46.8 ms to 3.2 ms. JSON remains the default, unchanged.

On the maintenance side, the team unified fuel/funds naming, with hydra-tui's ambiguous --cardano-signing-key renamed to --funds-signing-key. They also worked on several improvements, including a fix for a latent metrics bug in which mistyped Prometheus lookups silently failed, a correction to the published cost table that was underreporting minimum fees by roughly 30%, and a 90% reduction in the tx-cost-diff Nix closure size. Finally, they added a test to verify that partial fanout correctly recovers funds when an unburned token is present.

See Hydra’s technical report for more details.

RESEARCH

This week, the Research team is preparing to host their next session in a series of technical workshops under Cardano Vision 26 on data availability (DA). Registration is now open for the DA workshop on September 3, which aims to drive specifications and requirements, including preparing a Cardano Problem Statement (CPS). 

Additionally, the team had two papers accepted to Asiacrypt 2026. First, ‘A Tight Analysis of GHOST Consistency’ by Peter Gaži, Zahra Motaqy (UConn), and Alexander Russell. Second, ‘Crossing with Confidence: Formal Analysis and Model Checking of Blockchain Bridges’ by Pyrros Chaidos, Pooya Farshim, Denis Firsov, Dimitar Jetchev, Aggelos Kiayias, Markulf Kohlweiss, and Anca Nitulescu.


Weekly development report as of 2026-08-28

CORE TECHNOLOGY

This week, the ledger team continued advancing Leios support by extending the pool distribution to include the corresponding BLS key and per-pool BLS-attributed stake, making BLS-related pool information available wherever pool distribution data is consumed.

The team also began groundwork for a broader refactor of predicate failures toward standalone, reusable units shareable across rules and eras, with sub-transactions in Dijkstra as the initial target. They also relaxed CBOR decoders to accept indefinite-length encodings, making them tolerant of structurally equivalent CBOR variants.

On the testing side, the team extended sub-transaction test coverage, re-enabled several conformance test suites following generator fixes, and strengthened serialization tests in Dijkstra. See more details here.

LEIOS

Over the past few weeks, the Leios team completed the Earth phase of MusashiNet, the public testnet taking the Ouroboros Leios prototype to production-grade code. The phase ran 41 days from the June 23 launch, producing over 127,000 blocks and nearly 8,000 on-chain certificates, with pool participation growing from three to 63. Under synthetic load, the network peaked at 26.8 TxkB/s against the 4.51 TxkB/s the mainnet can reach with Praos alone – roughly six times current throughput, with parameters still at cautious initial values. The team shipped 12 releases in eight weeks.

Across the phase, the team hardened the prototype against the conditions only a live network exposes. They improved the node's database connection handling to eliminate a memory leak under sustained load, stabilized certificate validation so that syncing nodes reliably follow a chain carrying endorser blocks, and stopped the Haskell node re-encoding blocks on their way to disk, restoring byte-for-byte agreement across node implementations. Error handling was revised so that invalid input from an external peer results in a logged rejection rather than a halted process, and the mempool was redesigned to revalidate transactions off-chain, avoiding the lock the forge loop requires and keeping block production responsive as the mempool fills. All changes will go into testing during the Water phase, which is now open on a fresh chain.

The team is now finalizing a blog post covering the path to mainnet so far, with a detailed account of the challenges and their resolutions.

SMART CONTRACTS

The Cardano High Assurance team has been implementing the test tree, table results, and coverage views for the property-based testing extension, which will provide a brand-new user experience when using the testing tool.

The team also implemented a few more rules for the plu-stan static analyzer, bringing more automated checks for Plinth developers.

Finally, the team has been building new, more optimized formalizations of the cryptographic primitives for automated formal verification, enabling better usability and faster proofs. The team is also investigating several ideas to optimize Blaster to enable proof of larger scripts, paving the way for proof of several scripts as committed in the 26/27 Blaster roadmap.

DEVELOPER EXPERIENCE

This week, the developer experience team delivered a significant cardano-init update, roughly doubling the tool selection and introducing a shared protocol to ensure on-chain and off-chain tools work together consistently.

The team also added Plinth as an on-chain tool, Evolution SDK and Tx3 as off-chain options, and shipped a compatibility check that validates whether the selected tools and data provider work together – catching mismatches before project generation. Every scaffolded project now includes a stack-specific AGENTS.md file to guide both AI agents and developers through the project's interfaces and workflow. The CLI also received a full visual redesign with clearer error messages and improved diagnostics.

On the reliability side, the team added automated compliance tests for every template in the registry, fixed the doctor command to correctly recognize Tx3 projects, and merged the experimental-tools opt-in gate from last week.

The contracts-library continued at a steady pace with 15 commits, including a pull request to implement configuration parameter management. The Cardano Foundation's Developer Portal onboarding flow advanced with six further commits, and the team updated cardano-dev-skills – the knowledge layer that helps AI agents work within Cardano projects – adding support for Plinth, Tx3, and a new scalability skill.

Read the full update for more details.

SCALING

This week, the Hydra team delivered a notable performance improvement to the Blockfrost backend (#2821), replacing blind sleeps with condition-based polling. A full head lifecycle on preview dropped from over an hour to around 13.5 minutes, publishing from roughly 10 minutes to 32 seconds, and first seeding from around 30 minutes to 40 seconds. The team also fixed a crash under sustained load (#2819) caused by an uncaught error in the etcd gRPC link, and merged the rejectLowDeposits optimization (#2805), which is now a single-pass operation with a property test over multi-output UTXOs.

On the maintenance side, the team fixed a long-standing CI reliability issue (#2811): test port sentinels are now close-on-exec, preventing spawned subprocesses from inheriting bound ports – a bug responsible for 16 of the last 25 hydra-node CI failures. Flaky etcd, TUI, and chain-sync tests were also fixed, and withTempDir now cleans up atomically on success (#2813).

Finally, the team fixed some bugs and progressed work on benchmark reliability, opening several pull requests (#2827, #2828) to fix timing units, pin seeds, and improve end-to-end benchmark comparisons.

RESEARCH

This week, the research team is preparing for their next virtual technical workshop, details of which will be released soon.

Additionally, two research papers by Aggelos Kiayias have been presented at Crypto ‘26 in Santa Barbara, USA: Fast Difficulty Adjustment in Proof-of-Work Consensus by Juan Garay, Aggelos Kiayias, Yu Shen &

Permissionless Consensus from a Common Random String by Damiano Abram, Marshall Ball, Juan Garay, Aggelos Kiayias. New explainers are being written for these papers and will be shared soon.


Weekly development report as of 2026-08-21

CORE TECHNOLOGY

This week, the consensus team released prototype-2026w32 and continued advancing Leios prototype development. The node now generates and sends endorser-block announcements to peers, which is the first step toward full announcement-based diffusion. The team also added validation to reject malformed Dijkstra blocks and fixed several testnet observability issues, including the visualizer losing events and incorrect throughput dashboard units.

On node performance, the team removed allocations from transaction identifier comparisons – making them roughly five times faster – reduced allocations in mempool snapshot code, and simplified how the mempool moves transactions into a newer era.

The team released ouroboros-consensus 4.0.0.0 and 4.1, and enabled the node to hard fork into the next Dijkstra era. The default snapshot policy was updated to one snapshot per day on mainnet, keeping the node aligned with Mithril. Finally, they added a new documentation page covering hard forks and node-to-node versioning.

See this update for more details.

SMART CONTRACTS

Over the past two weeks, the Plutus team opened two CIPs and one CPS for community review:

CIP-0194 proposes adding a Match node to UPLC for matching on built-in constants, primarily Data. This would make consuming script contexts in validators significantly cheaper than the current approach. 

CIP-0195 specifies the Data encoding of the Plutus V4 ledger types. 

A new CPS aims to build community consensus on how to expose property tests and tests for older protocol and language versions to alternative node and UPLC evaluator implementers with minimal integration effort. Feedback on all three is welcome.

The team also merged the CollapseCase optimization pass, which rewrites list casing into dropList where applicable, updated the UPLC tool to deduce input and output formats from file extensions, and released Plutus V4 ledger API types in version 1.67.0.0.

DEVELOPER EXPERIENCE

This week, the developer experience team delivered 41 commits across four active deliverables, focusing on hardening and design.

The ecosystem map continued to serve as a living baseline following the D1 delivery, with significant curation around adding several new tools and stake pool explorers, including ODATANO, adapools.xyz, and Daedalus. OSS filtering was improved, and the developer pledge welcomed two new signatories.

On cardano-init, the team focused on making the tooling reliable and transparent under real conditions. They merged Yaci devnet CI coverage – the last scaffolding role without test coverage – and introduced a strict mode that converts silent skips into hard failures, ensuring a passing run genuinely proves that the devnet is booted and the full transaction round-trip is executed. They also introduced an experimental-tools gate, which flags tools that are not yet production-ready and prevents them from being scaffolded without explicit user confirmation.

The contracts-library saw 19 commits this week. The main contribution was a draft architecture for CIP-113 event-triggered assets – a design that allows tokens to respond automatically to real-world events through oracle data feeds, with built-in rules for how tokens can be transferred, upgraded, or retired. The team also expanded the library's catalog of contract explorations to include peer-to-peer swaps and prediction markets.

On the Cardano Foundation's Developer Portal, the team merged the beginner lecture track with improved lessons, theme-safe diagrams, and runnable Mesh examples. Building on this, intermediate lectures opened covering on-chain versus off-chain concepts, now also backed by runnable Aiken examples.

See the developer experience webpage for more details.

SCALING

This week, the Hydra team launched a new updates tracking webpage to improve visibility into ongoing work.

Across 9 merged PRs, 7 closed issues, and 48 commits, the team delivered progress across maintenance, performance, and ecosystem support. On the maintenance side, deposits too large to be claimed are now rejected, preventing funds from being locked on-chain with no path to redemption. They also updated Mithril to 2630.0 with the v2 database backend, fixed macOS release binaries to run without Nix, and resolved a Plutus time-conversion failure and a long-running-node era-history bug.

The team also merged the --deposit-activation flag, giving operators independent control over deposit activation timing. On the performance side, they extended the fanout benchmark to vary party counts and closed two PersistentQueue bugs affecting queue visibility and pop behavior. Documentation improvements from the previous week also landed, including the partial fanout page and several rendering and dark mode fixes.

See technical details.

RESEARCH

This week, the Research team is preparing the final version of the mid-year report to submit to Intersect. Additionally, two research papers by Aggelos Kiayias will be presented at Crypto ‘26 in Santa Barbara, USA, next week: Fast Difficulty Adjustment in Proof-of-Work Consensus by Juan Garay, Aggelos Kiayias, Yu Shen &

Permissionless Consensus from a Common Random String by Damiano Abram, Marshall Ball, Juan Garay, Aggelos Kiayias.


Weekly development report as of 2026-08-14

CORE TECHNOLOGY

This week, the performance and tracing team completed benchmarks for the Van Rossem hard fork, covering new Plutus execution budgets, internals, and a PlutusV3 baseline shift. The team also finalized the Hermod package restructure, added cardano-config support to trace-dispatcher, and delivered the genesis cache overhaul and protocol version 11 overlay.

On the tracing side, an alert manager feature for cardano-tracer is in the design phase, and native span support is underway in Hermod. For Leios, the team introduced a new beacon feature to impose memory limits, and benchmarking of on-disk LedgerDB transaction validation times is in progress.

See this technical update for more details.

The ledger team advanced Leios support by replacing placeholder types with an actual BLS key in pool parameters, and added a dedicated round-trip test suite covering Praos and Leios block-header serialization.

On nested transactions in Dijkstra, the team implemented the withdrawal mechanism for sub-transactions, reworked predicate failures to be more precise and consistent, and settled how UTXO state updates behave under nested transactions. The AdaPots obligations guardrail was updated accordingly, and a field for the initial account balance was added to the top-level transaction body.

On the Plutus side, the team introduced a per-block script cache so scripts decoded for one transaction can be reused within the same block, and added memoization for script hash computation.

Two CIPs advanced this cycle: the team implemented CIP-50 (pledge-leverage-based staking rewards) and began groundwork for CIP-23 by adding the minPoolMargin protocol parameter to the Dijkstra era. Test coverage was expanded by enabling the Conway conformance suite in Dijkstra and adding new tests around stake calculations and sub-transactions.

SMART CONTRACTS

The Cardano High Assurance team delivered a collection of optimizations for lean-blaster that resulted in a 30% performance improvement on smart contract benchmarks. They also added support for CVC5, which is currently in review. This aims to increase performance through parallel solving, access to additional theory solvers, and a standardized proof format for proof reconstruction.

The Visual Studio Code extension for property-based testing has also been made available for internal testing.

DEVELOPER EXPERIENCE

The developer experience team completed the Community Alignment deliverable (D1) ahead of the July 31 deadline. The milestone required a public map of the Cardano development ecosystem and a repository of key projects willing to collaborate on the shared roadmap – both are now in place.

The ecosystem map went live, offering graph, table, and analysis views backed by a curated dataset of Cardano tooling with identified gaps and overlaps. The signatory list for the developer pledge continued to grow, with several new names joining from across the ecosystem.

On cardano-init, the team merged a dependency-installer verification gate – a CI step that installs each declared dependency in a native environment and checks that binaries land on PATH. Standing up the gate also surfaced and fixed several broken installer recipes.

The contracts-library saw 17 commits as previously cataloged explorations continue to take shape as ready-to-audit contract code. The Cardano Foundation's Developer Portal onboarding flow also advanced with 28 commits.

With D1 complete, focus shifts to the cardano-init specifications and proof-of-concept milestone, due August 31.

SCALING

Over the past few weeks, the Hydra team released 2.3.0, headlined by a significant snapshot-processing speedup via the reapplyTx optimization, node configuration via a YAML file, native HD wallet key support, and several resilience fixes.

The team also worked on a major performance stack improving accumulator commitment computation – reducing a 4K-UTXO commitment from 2.13 seconds to 25 ms – alongside batched broadcast messages, raised transaction limits per snapshot from 100 to 1,000, and moved SQLite event encoding off the main node loop.

The team also implemented selective partial fanout, allowing operators to distribute a chosen subset of a closed head's UTXO rather than the full set. The deposit period is now embedded on-chain at initialization, a new --deposit-activation flag decouples activation from deadline calculation, and the commit endpoint now rejects deposits that could never be claimed.

Additionally, the team fixed a long-running issue where nodes rejected layer 2 Plutus transactions carrying a validity bound past the era forecast horizon, made hydra-chain-observer version-aware to remove compile-time coupling to a single protocol version, and fixed several usability issues in hydra-tui. Finally, they bumped Mithril to 2630.0 and switched to the v2 database backend.

RESEARCH

This week, the IO research team is gathering feedback from the community on the Cardano Vision ‘26 Mid-Year report, which is available for review on the Cardano Forum. The Cardano R\&D session held on July 16 also focused on updates from across the Cardano Vision ‘26 work streams. 

Additionally, members of the team recently attended the Science of Blockchain conference at Stanford University and the mini conference of the Stanford Input-Output Research Hub. The focus of this year was, among other topics, on censorship resistance, bridging Bitcoin to other chains, and the threats that quantum computers might present for deployed blockchains in the near future.


Weekly development report as of 2026-08-07

CORE TECHNOLOGY

Cardano's engineering is now decentralized. Since 2024, IO and its partners have been managing a transition that will conclude in 2027, resulting in several independent firms maintaining at least three Cardano node implementations in Haskell, Rust, and Go, alongside a complete set of formal specifications overseen by member-based organizations including Intersect and Pragma, and subject to community review and vote. As part of this process, IO is spinning out the Haskell node to community curation and control. Check out the engineering website for more details.

The ledger team continued integrating Leios features into the ledger, including block header abstractions and the addition of BLS12381 public keys to stake pool parameters and state, with the Dijkstra CDDL updated accordingly.

On governance and rules, the team implemented CIP-181, removing the DRep-delegation requirement for reward withdrawals, and introduced a safeguard to prevent accidental swapping of old and new Plutus cost models.

Work continued on making ledger interfaces more era-parametric, with expanded JSON serialization coverage and initial export and import modules for reading and writing canonical ledger state. Finally, they updated Conway conformance tests to protocol version 11, and added tests for the newly introduced top-level guard requirement.

See this update for more details. 

WALLETS AND SERVICES

This week, the Lace team released an update for iOS and Android, which introduces in-app swaps and restores the Banxa on-ramp on iOS. It also adds CardanoCube integration, and brings Ledger and Trezor hardware wallet support for Cardano. 

Users can now sort tokens by quantity, value, or ticker, add contacts or select their own accounts directly from the send flow, and access a new network information card in the Staking Center. The release also includes performance improvements and various refinements across the app.

DEVELOPER EXPERIENCE

This week, the developer experience team launched a new website for proposal tracking, which also provides weekly updates. Steady progress continues across all four active deliverables, with 27 commits landing this week.

On cardano-init, the CLI took a meaningful step forward with the addition of the fullstack option, which scaffolds a project that combines on-chain and off-chain code in a single setup – the starting point most teams use in practice. Work also began on multi-role framework support, including exploration of Scalus integration and a dedicated Scalus template, alongside early work on extending the doctor command to check dependencies outside a project context.

The contracts-library saw 12 commits this week, with development focused on feature branches, as the previously cataloged building blocks begin to take shape as real code. On the community side, the developer pledge is now live and open for signatures. The Cardano Foundation's Developer Portal onboarding flow also moved forward, with the scaffold for the new onboarding experience now in place.

Developers building on Cardano are encouraged to sign the pledge.

SCALING

This week, the Mithril team completed the refactoring of the non-recursive SNARK proof to remove the circuit verification key, upgraded the midnight-zk library to 2.3.3, enabled the Plutus-friendly hash for IVC proving, and refactored the error handling in the Midnight library. They also continued to analyze and optimize the testing strategy for the STM SNARK modules.

The team also continued implementing aggregator CLI commands for configuration parameters and creating a runbook for on-chain configuration parameters. They completed the simplification of the SignersRegistrationRetriever trait and continued work on shipping the Mithril signer node binary in the Cardano node bundle, the 2628 distribution release, the upgrade to Cardano node v.11.1, and the update of the DMQ node to 0.7.0.0.

Finally, they completed the repository transfer to the Intersect MBO organization and continued preparing the documentation website for Docusaurus v4.

VOLTAIRE

The Intersect treasury withdrawal governance action has been ratified, with DRep approval of 75.69% and Constitutional Committee endorsement by six of seven members. The withdrawal of ₳25.4M will fund technical stewardship and governance coordination for June 2026 to June 2027, with the transfer taking effect at the start of Epoch 646 (July 28).


Weekly development report as of 2026-07-24

CORE TECHNOLOGY

This week, the consensus team continued to stabilize the Leios testnet, releasing two prototype builds – prototype-2026w27a and prototype-2026w28 – and fixing several bugs identified during testnet runs. The team also implemented a voting dashboard to improve Leios observability, showing how endorser blocks are voted on and certified. They also reviewed a change to the endorser block certification, expected to roughly double the certification rate and increase throughput, and began drafting the Leios high-level design document.

Additionally, the team participated in the node diversity conformance testing meetings with the Amaru team and discussed how to specify the mempool and transaction submission precisely enough for other node implementations to follow.

They also improved the snapshot-converter tool's help and version output, making it clearer how to convert ledger snapshots between in-memory and on-disk formats (#2098), and added a reference document covering the metrics and tracing available for observing Ouroboros Genesis (#2103).

See this technical update for more details.

SMART CONTRACTS

This week, the Cardano High Assurance team delivered the Plinth Static Analyzer. It is available on the Visual Studio Marketplace – Plu-Stan (VS Code/Cursor). It comes with a set of binaries for an easy 1-click installation for most projects and architectures. It is still in its early phase, and all feedback is welcome.

The team also delivered the JSON-RPC implementation, which will serve as the backend for the property-based testing tool Visual Studio extension next quarter.

Finally, the team drafted the Universal Annotation Language specification, and several examples are being annotated with it. This will be used in surface languages (eg, Plinth) to directly annotate the business logic and security invariants of the various DApp smart contracts.

SCALING

This week, the Mithril team completed the modularity enhancement of the SNARK recursive circuit with gadgets. They continued work on refactoring the non-recursive SNARK proof to avoid carrying the circuit verification key, analyzing and optimizing the testing strategy for the STM SNARK modules, and upgrading the midnight-zk library to 2.3.3.

The team also enhanced the aggregator state machine to support the Blocked state and continued implementing the aggregator CLI commands for decentralized configuration parameters. Additionally, they continued shipping the Mithril signer node binary in the Cardano node bundle, working on the release of the 2628 distribution, upgrading to Cardano node v.11.1, and updating the DMQ node to 0.7.0.0.

Finally, the team completed the Rust upgrade to version 1.97, fixed the Mithril install script failing on the unstable version, and captured the end-to-end execution logs as artifacts.

VOLTAIRE

The van Rossem hard fork governance action met the required ratification thresholds across all voting groups on July 13 and was subsequently ratified at the epoch boundary of Epoch 643. According to protocol, the hard fork will automatically be enacted on the next epoch boundary on July 18, 2026 at 21:44:51 UTC (slot height: 192,844,800).

As well as Plutus improvements and Plutus Cost Model enhancements, this upgrade lays the foundation for the next upgrade, the Dijkstra era hard fork, which will introduce Ouroboros Leios to Cardano.

No other governance actions were directly impacted, delayed, or expired as a result of the hard fork governance action reaching ratification.

RESEARCH

This week, the research team released the Cardano Vision 2026

Mid-Year Progress & Transparency Report (Draft). This report is open for community feedback until Jul 30, 2026.

They also held their second Cardano R\&D session, during which the research leads shared updates on CV26 workstreams and the draft Mid-Year report. Watch out for the recording and video clips on our social channels next week. They also held a technical workshop on zero-knowledge (ZK) verification infrastructure for Cardano, which included presentations on Groth 16, Plonky 3, and Halo 2.


Weekly development report as of 2026-07-17

CORE TECHNOLOGY

This month, the performance and tracing team removed the legacy iohk-monitoring-based tracing backend from cardano-node, eliminating approximately 11,000 lines of outdated code across 15 modules and nine transitive build dependencies. The team also removed RTView from cardano-tracer, with its functionality now fully covered by the Grafana datasource and timeseries server. Work is underway to restructure trace-dispatcher into two distinct packages – hermod-tracing-api and hermod-tracing-core – allowing downstream packages to avoid pulling in the full backend stack when only instrumentation is needed.

The cardano-tracer timeseries HTTP API and accompanying Grafana datasource were merged, providing PromQL-like queries over metric timeseries. The team also moved the cardano-recon-framework and trace-resources package into the Hermod Tracing System project, and completed the genesis cache overhaul for the performance workbench – now close to merging.

Finally, benchmark optimizations showed a ~17% reduction in process CPU usage only under full saturation, with small increases of 3–6% in block adoption times. The default build remains recommended for the current Praos node.

The ledger team focused on several areas:

Performance: the StAnnTx memoization work was extended to Alonzo, completing integration across all relevant ledger eras. The long-running removal of the legacy Coders machinery also reached completion.

Leios support: the first Leios-related changes landed in the ledger, including the LeiosCert type added as an optional field to DijkstraBlockBody and a new cardano-protocol package to house protocol-level code.

Dijkstra era: work continued on ledger rule structural cleanup, including a new ENTITIES rule to handle withdrawals and direct deposits, and the third installment of Dijkstra conformance testing. The team also fixed two user-facing bugs – one affecting Plutus cost model overrides and one causing spurious TxInfo translation failures.

Tooling and infrastructure: the team moved generate-cbor CLI tool into cardano-ledger-api, added new Conway UTXO and script generators, and applied dependency updates across the stack.

WALLETS AND SERVICES 

This week, Lace released browser extension 2.1, one of its biggest updates yet. Headlining the release is wallet security check, a new feature that helps users identify whether a wallet may have been affected by a known security incident, either when restoring a wallet or by running a check at any time, with clear guidance provided where relevant. 

Alongside this, Lace 2.1 introduces a host of improvements across the wallet, including enhanced token management, streamlined send flows, expanded DRep ID support, improved network resilience, a smoother DApp experience, hardware wallet enhancements, and numerous stability and usability upgrades. The release also sets the stage for the upcoming Lace Carbon opt-in beta, giving the community an early look at what's next while helping shape the future of Lace through feedback.

SMART CONTRACTS 

Over the past weeks, the Plutus team has continued to make improvements across performance, tooling, and assurance.

Performance and compiler improvements. The team improved several parts of the Plutus optimization pipeline and ledger API implementation. This includes making common Value operations more efficient, adding a new compiler optimization for certain recursive functions, and fixing an optimization pass to preserve program behavior correctly.

Better benchmarking and tooling. They added lightweight benchmarking support to the UPLC executable, making it easier to sanity-check execution-time measurements for UPLC scripts and compare results against existing benchmark infrastructure.

Specs and formal methods. The team continued strengthening the foundations of Plutus by updating the UPLC specification for the built-in value type and related functions, and by improving the metatheory and certification infrastructure used to reason about compiler transformations. These changes make the formal-assurance workflow easier to maintain, debug, and scale to larger programs.

See these pull requests for more details.

DEVELOPER EXPERIENCE

The developer experience initiative officially kicked off last week, though work has been underway for several weeks now. Here is the current state:

A team of experts has been assembled across all deliverables

Work is ongoing in two of the three biggest workstreams: contracts-library and cardano-init; collaborators at the Cardano Foundation have significantly advanced the third workstream, optimizing onboarding on the Developer Portal, while the team finalized its contribution plan

contracts-library: the first reference contract is in place, with a set  of 20 reusable protocols and patterns for research and triage

cardano-init: specifications, a rough proof of concept, and a plugin architecture are ready for community validation – simple projects using Aiken, MeshJS, and infrastructure can already be created.

Have feedback on developer experience on Cardano? The team welcomes input – whether it's something that could work better or ideas for what to prioritize next. Reach out to Robertino via X, Discord, or email.

SCALING

This week, the Hydra team merged the user-friendly hydra-node configuration file (#2581, #2296), allowing operators to run a head from a single YAML file and compare configurations against peers at a glance. The team also scoped deposit and recovery chain observations to the head that produced them, closing a long-standing cross-head contamination issue (#2743, #2606). The Blockfrost backend now retries on HTTP errors instead of crashing and handles inline datums more reliably. 

Finally, the team fixed two issues in the event store's rotation logic – a race between VACUUM and streaming client history, and a deadlock in the network layer's outbound message queue that could silently halt broadcasting.

The Mithril team refactored the unsafe SNARK setup, prepared the prover input for the recursive SNARK aggregation primitives, decoupled the Midnight library in the SNARK recursive circuit, and tested that the degree of the SNARK circuits remains unchanged. They continued improving the modularity of the SNARK recursive circuit using gadgets.

The team also worked on enhancing the aggregator state machine to support the Blocked state and on capturing end-to-end execution logs in artifacts.

Finally, they continued work on shipping the Mithril signer node binary in the Cardano node bundle, releasing the 2628 distribution, upgrading the Cardano node to 11.1, updating the DMQ node to 0.7.0.0, and upgrading the midnight-zk library to 2.3.3.

VOLTAIRE 

Voting is proceeding in the constitutional committee election. DReps have until July 23 at 21:45 UTC to select their four preferred candidates from the ten people standing. See this X post for more details.

RESEARCH

This week, the research team is working on the Draft Mid-Year report for Cardano Vision 26. This mid-year report provides progress and transparency on all work streams. Links will be circulated soon on the Cardano Forum for community review.

The team also ran the first in a series of technical workshops on the Cardano Vision 26 programs, with this session focusing on zero-knowledge verification.


Weekly development report as of 2026-07-10

CORE TECHNOLOGY

This week, the consensus team focused on Mithril integration: nodes running on an on-disk backend can now export snapshots in the portable in-memory format, allowing Mithril signers to obtain the agreed format directly from the node. They also implemented a new snapshot policy that makes ledger state snapshot timing predictable using agreed-upon cadence values.

On the maintenance side, the team updated db-analyser to start analysis from a snapshot at a specified slot, avoiding full replays from genesis, and fixed a rare snapshot error on epoch-boundary blocks. 

Finally, the team shipped full-text search on the consensus documentation website, built as static files with no external service required.

WALLETS AND SERVICES

This week marked an important milestone for the Lace team as Lace Carbon entered internal testing.

Built on the foundations of Lace 2.0, Lace Carbon represents the next evolution of the wallet experience, with a focus on delivering a faster and more intuitive user journey.

Internal teams are now putting it through its paces and providing feedback ahead of a future opt-in release. We're looking forward to sharing more details in the coming weeks.

SCALING

This week, the Hydra team released 2.2.0, bundling the partial fanout implementation and hydra-tui revamp. The release also delivers a significantly expanded benchmark suite, including real-world TPS metrics, a Mixed UTXO generator, and a new matrix sub-command that sweeps cluster sizes and UTXO shapes. The team also unified on-chain membership checks around the BLS accumulator, reducing the snapshot-signing tuple and removing redundant hashing, and started optimizing snapshot processing so script-heavy heads avoid re-evaluating Plutus scripts that have already been validated.

The Mithril team completed the prover input preparation and SNARK proof verification for the recursive SNARK aggregation primitives and adapted the certificate chain to support recursive SNARKs. They continued work on creating the SNARK proof with the circuit, wiring the SNARK proof in the aggregate signature, refactoring the preparation of the prover input, adding the missing ancillary data for the IVC aggregate signature creation, and updating the Midnight ZK library audit status for SNARK.

Additionally, they completed the optimization of the end-to-end tests for SNARK, updated the ledger state snapshot selection in artifacts, upgraded to Pallas 1.1.0, and enhanced support for discontinued signed entity types in the aggregator.

Finally, the team continued work on shipping the Mithril signer node binary in the Cardano node bundle, the release of the 2625 distribution, and the fix for the aggregator switch in the explorer.

VOLTAIRE 

Candidate registration for constitutional committee elections closes this Sunday, June 21, at 21:45 UTC — and there is a development worth noting. Five candidates have now registered for four available seats, meaning the election is competitive, if only by a narrow margin. This is a meaningful improvement from last week, when the candidate count matched the number of open seats exactly.

Registration is still open. More candidates would strengthen the outcome, and anyone considering running should act now. Waiting until the last minute creates unnecessary risk, especially if there are wallet issues, signing issues, or questions about the registration process.

Once registration closes, DRep voting begins at the epoch boundary on June 23 and runs through July 23. The election uses stake-weighted voting: each DRep may vote for up to four candidates, and each candidate they vote for receives that DRep's full voting power. The four candidates with the highest total voting power at the close of voting will form the election outcome, which will then support an on-chain Update Committee governance action to seat the new members before current terms expire in September.

If you believe you have the experience, judgment, and commitment to contribute to Cardano governance at this level, please consider registering before the deadline:

Candidate registration

Election documentation

Weekly AMA sessions.

RESEARCH

This week, the research team is progressing work on Cardano Vision 26 workstreams and has published a blog on a recent Cryptographic Tools for Blockchains workshop at Eurocrypt '26.


Weekly development report as of 2026-06-19

SMART CONTRACTS

This week, the Cardano High Assurance team has finalized a beta version of the property-based testing tool and is pursuing internal testing. The team introduced a testing interface to easily explore a vast variety of scenarios, automated positive testing to verify the contracts meet their specifications, negative testing to ensure the testing interface is not overly restrictive, threat models to automatically discover common weaknesses, and testing coverage to ensure that no line of code was left untested.

The team is working on a Visual Studio code extension that will make exploring the testing outcome easier and debugging faster. They have also been experimenting with developing a complete AI skill that makes the process of writing this testing interface as simple as possible.

SCALING

This week, the Mithril team completed the implementation of the SNARK-friendly genesis certificate and the type-consistency and naming-alignment refactoring for the SNARK recursive circuit. They continued work on the prover input and the SNARK proof verification of the recursive SNARK aggregation primitives, and on adapting the certificate chain to support the recursive SNARK.

Additionally, they completed support for DMQ ledger peers, enforcement of the DMQ message ID format, and a prototype for Cardano node ledger state certification. They also continued shipping the Mithril signer node binary in the Cardano node bundle and updating the ledger state snapshot selection in artifacts.

Finally, the team completed the upgrade of the DMQ node to 0.6.0.0, removed support for Mithril nodes in the devnet, and continued work on optimizing the end-to-end tests for SNARK.

VOLTAIRE 

The call is out for more candidates to register for the Cardano constitutional committee. Four seats will become vacant as members’ terms expire in September, and only four candidates have registered for the upcoming election. If no additional candidates register before the June 21 deadline, there will be no competitive election.

This committee is very important to Cardano governance because many governance actions need approval from two-thirds of the constitutional committee.

Several people have already indicated their interest. If you feel you can contribute to this important committee, register today.

RESEARCH

This week, the research team's IOR proposal for Cardano Vision 2026 was ratified with a confidence vote of 74.96%. Thank you to all the DReps and the wider Cardano community who voted in favor - your support and confidence are appreciated. Community feedback across all votes is valued and will inform delivery going forward.


Weekly development report as of 2026-06-12

CORE TECHNOLOGY

This week, the performance and tracing team published release benchmarks for node v.11.0.1 and ran feature benchmarks for TxSubmissionLogicV2 and compiler version comparisons. On the development side, the team removed legacy tracing and completed work on genesis caching and post-processing infrastructure. The cardano-tracer HTTP API for metric time-series queries and Grafana datasource integration is also complete and awaiting merge. Leios mempool benchmarks using tx-centrifuge are underway. Finally, the team merged the formal trace schema definition with the conformance framework to be presented at the Porto workshop. Learn more here.

The consensus team worked on Leios prototype development, starting the migration to the Dijkstra era of the Cardano node, enabling stake pool operator key registration, and aligning the prototype with the audited BLS cryptography. Work is ongoing on endorser block application, mempool partitioning, and throughput dashboard integration.

On fetch and network design, the team continued investigating multi-peer endorser block fetch strategies, refined the safety argument for Linear Leios, and concluded that erasure coding is unlikely to pay off at this stage. The change to the committee selection scheme was merged into the Leios SIP, reducing the certificate size from around 8 kB to 200 bytes.

The team also began discussions on mempool design for Leios, with approaches under consideration including two distinct mempools and a directed acyclic graph for the ledger.

As part of the UTXO-HD and Ledger-HD initiatives, work continued on the UTXO-HD redesign, LSM snapshot export, and the migration of InstantStake to the redesigned UTXO-HD as a pattern for the remaining datasets. Finally, the team worked on predictable ledger state snapshots in cardano-node and on enabling search on the consensus documentation site.

WALLETS AND SERVICES

This week, the Lace team released Lace 2.0.6. This release resolves the issue with DRep ID message signing on mnemonic wallets, enabling authentication for the Intersect Hydra voting DApp. DRep IDs are now displayed in CIP-129 format for improved clarity and compatibility, stake key message signing is supported on Ledger hardware wallets, and connected accounts are always shown during message signing flows. The release also fixes an issue where network selection was not always preserved when migrating from Lace v1 to v2.

SCALING

This week, the Hydra team merged the partial fanout feature and is preparing for the release. Heads with large UTXO sets can now be fanned out in multiple steps, each verified on-chain via a BLS accumulator membership proof, with the chunk size dynamically determined by binary search to maximize outputs per step. This removes the previous limit on UTXOs per head. Alongside this, the hydra-tui received a significant refactor, the test suite migrated to tasty, and several correctness fixes landed around cross-head event contamination and recovered deposits leaking back into L2.

The Mithril team completed caching of circuit keys for the SNARK circuit in the STM library and removed the helpers module for the SNARK recursive circuit. They continued work on the type consistency and naming alignment refactoring of the SNARK recursive circuit, the prover input of the recursive SNARK aggregation primitives, and the implementation of the SNARK-friendly genesis certificate.

The team also completed enhancements to support the genesis verification key in the explorer. They continued work on shipping the Mithril signer node binary in the Cardano node bundle, enforcing the DMQ message ID format, prototyping the Cardano node ledger state certification, and removing Mithril node support in mithril-devnet.

Finally, the team worked on optimizing end-to-end tests for SNARK and fixing flakiness in the CI.

RESEARCH

This week, the research team is sharing content from last week's Cardano R\&D session, which delved deeper into the IOR proposal, focusing on post-quantum security, exploratory research on sharding, and updates on consensus, scalability, and other workstreams in Cardano Vision 2026. If you missed the event or want to catch up on discussions, you can watch the recording. 

The team has also published a blog post on Linear Leios outlining the Leios roadmap to solve the blockchain trilemma.

The IOR proposal future-proofs Cardano. We are down to the final days – make your vote count today!  Voting is open until June 8.


Weekly development report as of 2026-06-05

CORE TECHNOLOGY

This week, the ledger team focused on the Plutus context performance work: the validation rules now consume precomputed information stored in StAnnTx rather than recomputing it, completing an optimization arc spanning several cycles. In parallel, work began on removing the legacy Coders machinery from the codebase, with several call sites already migrated to a new SparseKeyed alternative.

On the Dijkstra era front, the team introduced a dedicated transaction\_mempool rule to distinguish the mempool transaction decoder from the one used inside the block body, tightened the chain\_code specification to 32 bytes, and completed a fixes pass ensuring all CDDL tests pass. 

Conformance testing preparation for Dijkstra also advanced, and the canonical ledger state work continued with a new entities/accounts/v0 namespace and a refactor of gov/proposals/v0 to preserve proposal order.

Finally, the team added two new microbenchmarks for the TICK and NEWEPOCH rules to support epoch transition performance analysis, and introduced a new generate-cbor CLI tool for generating CBOR for any rule in any era from the command line – useful both internally and for downstream consumers such as serialization libraries and alternative node implementations.

SMART CONTRACTS

This week, the High Assurance team completed formalizing the latest built-ins, arrays, and values, and achieved 100% compliance with the conformance test suite. The tool is ready for the upcoming hard fork changes and will run on the latest release of Plutus core.

Last but not least, the community granted funding to the consortium between Harmonic Labs, IO, Lantr, Midgard Labs, No.Witness Labs, SAIB, and TxPipe to develop an extended version of Blaster, integrate it into Aiken, Pebble, Futura, and Scalus, and improve the user experience through a container-based development environment, Visual Studio Code extensions, and generic property sets and security checks.

SCALING

This week, the Mithril team completed the analysis of the impact of the recursive SNARK on the security of the Mithril protocol, implemented benchmarks for the non-recursive SNARK, and refactored the error handling of the SNARK recursive circuit. They continued work on circuit key caching for the SNARK circuit in the STM library, removing the helpers module for the SNARK recursive circuit, the prover input for the recursive SNARK aggregation primitives, and implementing the SNARK-friendly genesis certificate.

They also enhanced the synchronization of immutable files in the Cardano database, implemented robust support for unknown and in-progress signed entity types, removed the Cardano database v1 backend, and continued working on the prototype for Cardano node ledger state certification. Additionally, they enforced the DMQ message ID format and enhanced the support of the genesis verification key in the explorer.

Finally, the team completed the update of the protocol security page on the website and continued work on shipping the Mithril signer node binary in the Cardano node bundle.

The Leios team made progress across protocol design, implementation, and testing.

The Cardano community voted to approve Leios' funding through the treasury with 88% support, and the on-chain action is poised for enactment. This secures the resources needed to carry Leios from prototype to mainnet hard fork.

On the protocol design side, the team merged an update to CIP-0164 that replaces the original voting committee design with a stake-based committee selection scheme. The practical effect is a significant reduction in certificate size – from ~8 kB to~ 200 bytes, roughly 40× smaller – reducing bandwidth, storage, and validation cost on every block.

The Leios reference implementation was successfully rebased onto Dijkstra, the upcoming Cardano ledger era, in a cross-team effort spanning the ledger, network, consensus, and node repositories.

Finally, the dedicated red team demonstrated Piranha, a new attack-simulation tool designed to coordinate and evaluate attacks on the protocol, supporting the high-confidence testing track.

VOLTAIRE

Cardano’s constitutional committee elections are coming, with candidate registration now open, and voting scheduled from June 13 to July 13. As usual, Intersect has stepped in to represent the community in facilitating this election, even though it is a Cardano election, not an Intersect one.

Candidate registration closes on Saturday, May 30, so there’s still time to go to the voting page and submit your application. However, note that, unlike in previous years, when generating credentials before registering was optional, this year it is mandatory.

RESEARCH

This week, the research team hosted the first Cardano R\&D session of 2026, in which researchers outlined the workstreams of the IOR proposal that will shape Cardano’s development. They discussed post-quantum security, Ouroboros Leios, ZK capabilities for Cardano, and explored examples of foundational research such as sharding, incentives, and relationship credentials. 

Voting on the IOR proposal is open for DReps until June 8.


Weekly development report as of 2026-05-29

CORE TECHNOLOGY

This week, the consensus team delivered the first voting capability in the Leios prototype: nodes now diffuse votes over a dedicated mini-protocol, and a voting thread casts votes on completed endorser-block closures. This forms the foundation for committee-based endorsement and is exercised by new threadnet property tests (#1963).

The team continued working on Leios, reworking the prototype branch to target the same ouroboros-consensus-3.0.1.0 release delivered as part of cardano-node v.11.0.1, so downstream consumers building against that node release can pick up Leios without a separate consensus branch (#2041). The team is also adding late-join support, so a node joining the network after an endorser block was produced can still resolve the resulting certified blocks (#2040), and replacing the placeholder voting from #1963 with stake-based committee selection and real BLS signatures, so votes are individually validated before being relayed (#2039). Performance work on the in-memory Leios database is also underway, addressing contention and laziness issues that were causing nodes to time out under load (#2032).

As part of treasury funding initiative 10, the team retired the V1 LedgerDB implementation and the LMDB backing store. V2 has been the default for some time; removing V1 deletes a substantial amount of unreachable code, drops the LMDB dependency, and simplifies the LedgerDB API – snapshots no longer block the caller, and the tryFlush no-op has been removed (#2030). This paves the way for adding more tables to the ledger state and storing them on disk.

WALLETS AND SERVICES

Lace is now available on iOS, completing the multi-platform rollout launched with Lace 2.0. One wallet, three chains, available on browser and mobile.

SCALING

The Mithril team completed the refactoring of the recursive circuit, the preparation of the prover input implementation in the STM library, the off-circuit verification tests for the recursive SNARK circuit prototype, and the replacement of the temporary certificate circuit with the STM circuit. They also continued work on circuit key caching for the SNARK circuit in the STM library, the recursive SNARK aggregation primitives prover input, the preparation of the SNARK-friendly genesis certificate implementation, and the non-recursive certificate circuit benchmarks.

Additionally, the team continued work on shipping the Mithril signer node binary in the Cardano node bundle, robust support for unknown and in-progress signed entity types, enforcement of the DMQ message ID format, and enhancements to immutable file synchronization for the Cardano database.

Finally, the team completed enforcement of Mithril crate versions in downstream Mithril crates and added more details to the protocol security page on the website.

RESEARCH

This week, the research team hosted an X Space featuring Charles Hoskinson and Professor Aggelos Kiayias to dive into the IOR proposal. The discussion centered on the three themes of the proposal: post-quantum security, scalability, and human-centered design. The speakers emphasized that Cardano’s first-principles methodology is the soul of the brand. This approach has been vindicated, as other ecosystems are pivoting towards the same peer-reviewed, academic, research-based foundations that Cardano pioneered.

Ultimately, the session underscored that a positive vote on this pivotal proposal is an investment in future-proofing Cardano, ensuring it retains its science-based advantage and remains the industry leader in evidence-based blockchain development.


Weekly development report as of 2026-05-22

CORE TECHNOLOGY

This week, IO’s core technology teams released Cardano node v.11.0.1, the first release to support the protocol version 11 intra-era hard fork. This enables the upgrade to protocol version 11 once the hard fork governance action has been voted on by stake pool operators, delegated representatives, and the constitutional committee, and enacted on the chain. The release also bumps cardano-api and cardano-cli to their 11.0 series and advances the experimental hard fork target to protocol version 12. Additionally, it adds HTTPS support for EKG and Prometheus metric servers in cardano-tracer, along with a new timeseries store exposed via REST API.

Benchmarking reports relevant to the 11.0.1 release are available in this post.

The ledger team continued preparations for the upcoming Dijkstra era.

On CIP-118 nested transactions, the team updated the CDDL specification of block to match CIP-0176 non-segregated block body serialization, added a dedicated block\_body, and added an EncCBORGroup instance for DijkstraBlockBody to unblock downstream integration in consensus. They also fixed serialization for DijkstraTx and integrated StAnnTx – the state-annotated transaction representation that memoizes the Plutus parts of validation – into the LEDGER, UTXOW, UTXO, and MEMPOOL rules across every era, realizing the performance benefits of earlier memoization work end to end. Several guards were also added to the Plutus V1–V3 context translation so that new Dijkstra features unsupported by older language versions cause translation to fail explicitly rather than silently producing an incorrect context. The team also completed preparation work for running conformance tests in Dijkstra and began reorganizing the test hierarchy around a new class-based ImpSpec approach that enforces a common set of tests across all eras.

On the Ledger-HD, the streaming infrastructure for injecting large volumes of data from genesis files is now complete, covering stake pools and the remaining state needed for testing and benchmarking work targeting constant-memory reads from disk.

The CIP-165 canonical ledger state expansion continued with two additional namespaces added under the new namespace-based query interface: entities/dreps/v0 and entities/stake\_pools/vrf\_key\_hashes/v0.

Finally, on the maintenance side, the team expanded examples and added golden tests for all queries migrated from consensus to ledger, and did the same for transaction examples. They published the changelog for the cardano-node-11.0 release, made ZeroTreasuryWithdrawals a permanent governance check for every era, starting with Conway, and switched ChainData and Attributes to use ByteArray instead of ByteString to avoid pinned-memory fragmentation. Finally, they adjusted several NoThunks instances and fixed a cuddle failure on Plutus Constr data and a bug in its CDDL spec definition.

SMART CONTRACTS

This week, the Plutus team published a blog post about the improved UPLC command-line tool. Among other things, the tool can optimize Plutus scripts, with experiments on recent mainnet scripts indicating average execution cost savings of more than 10%.

Several further improvements to UPLC optimization have been integrated into the tool in recent weeks, including a new --inline-unconditional-growth flag for tuning inliner aggressiveness, Agda certifiers for several optimization passes, a new constant-folding pass, and a pass that hoists polymorphic built-ins to reduce the use of force. Common subexpression elimination (CSE) has also been improved.

Feedback, bug reports, and ideas for further improvements are welcome via GitHub issues in the Plutus repository.

SCALING

This week, the Hydra team released hydra-node 2.1.0, bringing improved deposit security, a new SQLite-backed event store, a ~7% reduction in snapshot confirmation latency, and compatibility with the Cardano node protocol version 12+. All operators are encouraged to upgrade. 

The team also upgraded to cardano-api 11.1 and cardano-node 11.0.1 to support protocol version 12+, and switched from a lazy to a strict Map to reduce memory pressure during transaction processing. Work continues on partial fanout and a user-friendly configuration file for hydra-node to replace the current command-line flags.

The Mithril team completed the trusted setup validation for the SNARK circuit in the STM library, the implementation of standard Schnorr signatures in the STM library, the implementation of the SNARK-friendly protocol message, the in-circuit and off-circuit verification tests for the recursive SNARK circuit prototype, the preparation of the recursive circuit refactoring, and the preparation of the prover input implementation in the STM library.

The team also prepared the ledger state certification proof of concept and continued work on enforcing the DMQ message ID format.

Finally, they completed the upgrade to Cardano node v.11.0, updated all GitHub runners to Ubuntu 24.04, and continued work on robust support for unknown and in-progress signed entity types.

RESEARCH

This week, the research team held a session on Cardano Over Coffee to discuss the IOR treasury proposal, hosted by Fergie Miller, director of research partnerships. The team is also hosting an X Space next week with Aggelos Kiayias, chief scientist at IO. Details will be shared on our socials soon!


Weekly development report as of 2026-05-15

CORE TECHNOLOGY

Over the past few weeks, the consensus team has continued working on the Leios prototype development as part of the treasury funding initiative 4. The team completed migrating endorser-block announcements and certification flags from the ledger block to the Praos header, aligning the prototype with the original Leios design (#1978, ouroboros-leios#874, cardano-node#6537). They also fixed several bugs in the Leios prototype, helped the system reliability team spin up a Leios testnet that is now running the prototype, and continued working on the voting prototype as well as on the fix for in-memory LeiosDB performance issues observed during the March demo (ouroboros-leios#844, cardano-node#6554). The team is now working on applying certified endorser blocks to the mempool. 

Additionally, as part of the Ledger-HD (treasury funding initiative 11), the team refactored consensus-type variables so that ledger tables, transaction inputs, and transaction outputs are indexed by the block type rather than the ledger-state shape, collapsing duplicate type-class instances and simplifying the codebase ahead of Ledger-HD integration (#2016, #2019).

In terms of maintenance and support (treasury funding initiative 17), the team pre-released Cardano node v.11.0 and expanded the documentation by adding an explainer on the consensus layer's ticking mechanism. 

SCALING

This week, the Leios team merged the documentation website and development tracker into a unified site. Additionally, in their April session, project leads discussed a successful prototype demonstration that achieved significant transaction throughput while testing new features including vote diffusion and included several fixes. The team reported progress on benchmarking and stress testing efforts aimed at optimizing transaction validation and network efficiency across various hardware configurations. They also made technical adjustments to committee selection and cryptographic keys, moving toward more deterministic and smaller certificates for better scalability. Additionally, the team progressed work on network fetching logic and simulation models to ensure robustness against latency and adversarial conditions, and reviewed the development roadmap covering the progression from controlled testing environments toward a production-ready mainnet release.

The Hydra team made progress on partial fanout, including KZG trusted setup with G1 accumulator commitments. They switched from list to Seq for faster benchmark runs at higher transaction counts and fixed a file-handle exhaustion issue by ensuring connections are closed properly. They also upgraded Mithril compatibility, added a fallback so nix develop works on aarch64-darwin, and updated the Pumba flake to build on modern Mac hardware.

The Mithril team completed the data encoding and state transition tests for the recursive SNARK circuit prototype, as well as the detection of SNARK circuit modifications in the STM library. They continued work on implementing a SNARK-friendly protocol message, the trusted setup validation for the SNARK circuit in the STM library, the implementation of Schnorr signatures in the STM library, and the in-circuit verification tests for the recursive SNARK circuit prototype.

The team also continued work on robust support for unknown and in-progress signed entity types, as well as enforcement of the DMQ message ID format. They completed the ledger state snapshot conversion tests in end-to-end tests and added support for stable and unstable versions of the OpenAPI specification.

Finally, they accelerated SNARK circuit tests in CI, upgraded Node.js 20 GitHub Actions, and fixed flaky nightly tests and circuit golden tests.

RESEARCH

This week, the research team has submitted their proposal on-chain and published a blog on research as one of Cardano’s core competitive advantages.


Weekly development report as of 2026-05-08

CORE TECHNOLOGY

IO's treasury proposals are live on the proposals website. For context on previously funded work, the team published the Q4 2025 – Q1 2026 delivery report, which covers what IO delivered for Cardano under the previous funding cycle.

To support the community in reviewing the proposals, IO has been hosting a series of X Spaces. A general kickoff session launched the process, followed by two dedicated spaces covering four proposals: Cardano upgrades, developer experience, Cardano High Assurance, and Plutus. The April 29 session covered Leios and layer 2 scalability proposals. Follow IO on X for more details on the upcoming spaces.

Delegated representative (DRep) voting is open now and closes on May 24. Explore the full set of proposals.

The ledger team progressed work on ledger rules for nested transactions and on translating newer features into the Plutus context. They introduced a change to how the Plutus context is prepared for each transaction entering the mempool – when fully integrated, this is expected to result in meaningful performance improvements.

Additionally, the team transferred all ledger state queries from the consensus component to the ledger component and introduced an initial stage of streaming injection functionality. This will be used for testing and benchmarking, particularly for on-disk storage scenarios where large volumes of data need to be injected into the ledger state without affecting operating memory.

Finally, they fixed several CDDL bugs identified using the new AntiGen and cuddle tools, and improved examples and facilities for golden tests.

WALLETS AND SERVICES

Lace mobile is now available on Android, with iOS support to follow. The release brings core wallet functionality to mobile – sending and receiving assets, staking, swapping, and DApp connectivity – without requiring a desktop. This release also marks the start of Lace 2.0, a step toward a multi-chain experience across Cardano and Bitcoin.

SMART CONTRACTS

This week, the Cardano High Assurance team demonstrated two new tools that together enable automatic correctness proofs for Cardano smart contracts, from source language down to the CEK machine, against the actual ledger rules.

The first, PlutusCoreBlaster, is a complete executable formalization of the Cardano CEK machine, UPLC, and Plutus Core built-ins in Lean 4. Any UPLC program can be imported, and correctness theorems can be written against its actual execution – not a model, but the real implementation. This means developers can use the UPLC output from any surface language (Aiken, Plinth, Plutarch, or others) without needing to model their contract separately.

The second, CardanoLedgerAPIBlaster, is an executable formalization of V1, V2, and V3 script contexts and ledger rules, reflecting the actual design choices made by the node – consistent with what runs on preview, pre-production, and mainnet.

To demonstrate what the tooling unlocks, the team formalized the sellNFT contract using CardanoLedgerAPIBlaster and ran Blaster on it. The proof revealed a double-satisfaction vulnerability. Combined, the two tools give developers a complete path: write a contract, import it, state your properties, and Blaster automatically proves them against the real CEK machine and ledger rules.

Join the team's Discord discussions and see the 2026 funding proposal to extend this work on the proposals website.

SCALING

Over the past few weeks, the Leios team improved the prototype's storage system, which had been affecting the test suite and proto-devnet runs, restoring stability in both environments. They published the first ledger throughput benchmarks for processing transactions with and without validation, confirming the expected speedup from reapplication over initial validation. More coverage will be needed, but the team is cautiously optimistic about this crucial aspect that underpins the Leios protocol's security. 

The team also began work on replacing the current 'always certified' shortcut with real vote traffic, while also aiming to simplify the committee selection scheme, allowing for smaller votes, certificates, and faster verification. Additionally, they updated the prototype to place endorser-block announcements in the Praos block header, aligning the implementation with the protocol specification. The team updated the cost estimate analysis and interactive tool to use the CIP-164 Linear Leios protocol, providing better guidance on expected operational costs for SPOs.

Finally, the liveness proof for the Agda formal specification – covering honest chain growth and chain quality – was submitted for review, closing the last major open gap in the formal specification.

The Mithril team released the 2617.0 distribution – read the developer blog for more details. This version bumps the Mithril signer to 1.0.0, marking it as production-ready on the release-mainnet network, adds support for Cardano node v.10.7.1, and includes statically built binaries as well as various bug fixes and improvements.

The team also completed the golden tests for the recursive SNARK circuit prototype and enhanced witness preparation for the non-recursive prover in the STM library. They continued work on implementing a SNARK-friendly protocol message, the data encoding and state-transition tests for the recursive SNARK circuit prototype, and detecting SNARK circuit modifications in the STM library. Additionally, they completed the extraction of the Merkle tree modules into a new crate.

Finally, the team implemented support for multiple tiers of tests in CI, refactored the crates publication, and investigated test flakiness.

VOLTAIRE

Voting in the Intersect election committees concluded on Friday, May 1, at 12 noon UTC. With 105 candidates for 37 available seats, the election was hotly contested. An independent audit has commenced and will continue over the weekend. Results will be announced on Monday, May 4, at noon UTC.

RESEARCH

This week, the research team has submitted their treasury proposal and are preparing their DRep Advocacy campaign, which includes content and events to explain the impact and value of the workstreams in the IOR proposal.


Weekly development report as of 2026-05-01

CORE TECHNOLOGY

This week, IO’s treasury proposals went live on the proposals website. The teams submitted nine proposals focusing on performance, scalability, and capability improvements, with a funding ask of just under 50% of last year’s and a clear focus on delivery.

Delegated representative (DRep) voting is open now and closes on May 24.

Explore the full set of proposals.

This week, the consensus team continued Leios prototype development as part of the 2025 Treasury Funding Initiative 4 (Leios implementation). They fixed a crash in the prototype’s fetch decision logic (#1996), added cumulative transaction bytes to the ledger state and exposed them as a Prometheus metric – giving operators a direct measure of throughput on the proto-devnet (#1988, cardano-node#6536) – and added a confirmed-transaction throughput panel to the proto-devnet dashboard (ouroboros-leios#859). They also hardened the prototype’s test suite around endorser-block inclusion to make it easier to extend to upcoming features such as voting (#1971), and started aligning the prototype with the original Leios design by migrating the endorser-block announcement and certification flag from the ledger block into the Praos header (ouroboros-leios#837). Additionally, the team is investigating a performance bug observed during the March demo, in which the time between forging and storing an endorser block can stretch to minutes, disrupting protocol operation (ouroboros-leios#844). A work-in-progress snapshot of the SQLite latency investigation from March was shared, which informs Leios-related storage work (ouroboros-leios#863), and the logbook update from the 2026-04-09 consensus office hours was posted (ouroboros-leios#854).

As part of the 2025 Treasury Funding Initiative 17 (maintenance and support), the team defined sensible default values for PerasParams, simplifying configuration for downstream users (#1998). On the documentation side, they added an explanation page describing the data flow between consensus components (#1955), published a first iteration of an explanation page covering the Ouroboros Praos consensus protocol (#1960), and removed leftover documentation stubs while cleaning up the sidebar on the consensus website (#1956). Finally, they released ouroboros-consensus-3.0.1.0 (#1987), bumped the node’s consensus dependency to 3.0.1 to propagate recent improvements to cardano-node (cardano-node#6533), and re-landed the rework of snapshotFromIS as part of ongoing improvements to ledger snapshotting (#1992).

SMART CONTRACTS

This week, the Plutus team made several improvements to the UPLC executable. They added support for hex-encoded scripts and blueprint JSON as inputs, with the ability to optimize all scripts they contain. They also added the ability to certify optimizations and generate a certifier report for each optimized validator, addressed several flags to give users control over optimization behavior, and enabled evaluation of a script with user-supplied input before and after optimization.

The UPLC executable is included with every Plutus release. While it supports a range of use cases, its most immediately useful feature for most developers is script optimization. Preliminary testing suggests that applying these optimizations to recent mainnet scripts yields, on average, a 10% reduction in execution cost and a 2% reduction in script size. User documentation for the UPLC executable is currently in progress.

The Cardano High Assurance team delivered a completely formalized example from the Invariant0 Capture-the-Flag, showing how users can formalize their own smart contracts and discover vulnerabilities.

SCALING

This week, the Mithril team completed production readiness preparations for circuit keys and trusted setup in the STM library. They continued work on the golden tests for the recursive SNARK circuit prototype, enhancements to witness preparation for the non-recursive prover, deterministic proof generation in golden tests, and data encoding and state transition tests for the recursive SNARK circuit prototype.

The team also completed the explorer update for Cardano blocks and transactions, the documentation of the Cardano blocks and transactions signature on the website, and the plan to sunset Cardano transaction certification. They continued work on the upgrade to Cardano node v.10.7 and on the preparation of the next distribution, 2617.

Finally, they investigated an issue with signer registration in the SNARK prover, fixed flaky end-to-end tests, and worked on supporting multiple tiers for CI tests.

VOLTAIRE

In the latest news on the Intersect committee elections, 105 candidates are standing for the 37 available seats.

Voting is underway and continues until 12 noon UTC on Friday, May 1.

RESEARCH

This week, the research team is working on their treasury proposal.

Read more about the team’s strategic collaboration with the Archimedes Research Unit in Greece in the latest blog post.


217 Search Results

Weekly development report as of 2026-09-25

Weekly development report as of 2026-09-18

Weekly development report as of 2026-09-11