Canton devkit (dpm component) - a unified toolkit for Canton app development (LocalNet, dar helpers, etc) #18 Milestone 2: Web UI, Observability, Monitoring, DAR & Contract Tooling, Optional AI Agent Skill Documents

[pedrodneves]

Milestone 2: Web UI, Observability, Monitoring, DAR & Contract Tooling, Optional AI Agent Skill Documents

• Estimated Delivery: Month 6
• Focus: Web UI for LocalNet management, integrated observability, DAR package management, live contract and transaction exploration, and optional AI agent skill documents.
• Deliverables / Value Metrics:
  • Web UI covering all CLI features from Milestone 1 with a user-friendly interface.
  • Richer LocalNet automation conveniences, such as machine-readable status output, environment export for app/test configuration, named-instance discovery, enriched doctor diagnostics, and deeper troubleshooting guidance.
  • Example CI workflow demonstrating LocalNet startup, readiness wait, optional DAR upload, test execution, and teardown.
  • Bundled Prometheus/Grafana stack with per-component enable/disable, sensible lightweight defaults, and documentation of minimum practical resources when the full stack is enabled.
  • Canton-specific Grafana dashboard presets focused on DApp developers: transactions/sec, command completion latency, active contract counts, and per-template throughput.
  • dpm localnet metrics subcommand printing Grafana dashboard URLs and a concise text summary of key metrics (throughput, latency p50/p99, resource usage).
  • DAR management CLI (dpm localnet dar upload/list/info/download/diff/remove/build-upload/watch) with multi-participant support, optional dpm build integration, and SCU-aware diff signals.
  • DAR Web UI with drag-and-drop upload, per-participant vetting toggles, package explorer tree, diff viewer, and hot-deploy indicator.
  • Contract tracking CLI (dpm localnet contracts watch and dpm localnet tx ls/replay) backed by Ledger API v2, scoped to capabilities not already provided by dpm daml-shell (live watch, multi-filter transaction listing, per-party visibility projection).
  • Contract tracking Web UI "Explorer" with live ACS table, transaction timeline, contract detail drawer, and explicit per-party visibility projection.
  • Optional AI agent skill documents demonstrating safe dpm localnet workflows for LocalNet lifecycle, DAR upload, package inspection, contract queries, and log/status checks.
  • Documentation on recommended usage, dashboard customization, DAR workflows, contract explorer usage, and optional AI agent skill documents.
• Adoption Metrics: at least 5 companies/teams have started using it in their daily Canton development workflow.

Milestone	Target deadline	Funding Request
Milestone 2 (Web UI, Observability, Monitoring, DAR & Contract Tooling, Optional AI Agent Skill Documents)	Month 6	400,000 CC upon committee acceptance.

Originally posted by @pedrodneves in #18 (comment)

[zheli]

Milestone 2 Delivery — Canton DevKit (Web UI, Observability, DAR & Contract Tooling, AI Agent Skills)

Hi team, happy to report that Milestone 2 is complete and ready for review.
Full accounting against each acceptance criterion with evidence below.

Repository

https://github.com/bitdynamics-ab/canton-devkit

Acceptance Criteria Checklist

Deliverable	Evidence
Web UI covering all M1 CLI features, user-friendly interface	frontend/src/screens/ (Dashboard, InstanceDetail, CreateLocalNetModal, DoctorScreen, BackupRestore, etc.)
Richer automation: machine-readable status, env export, named-instance discovery, enriched doctor, deeper troubleshooting	internal/cli/localnet/status.go, internal/cli/localnet/env.go, internal/cli/localnet/list.go, internal/cli/localnet/doctor.go; docs: docs/troubleshooting.md
Example CI workflow (startup, readiness wait, optional DAR upload, test run, teardown)	examples/ci/github-actions.yml, examples/ci/gitlab-ci.yml, examples/ci/README.md;
Bundled Prometheus/Grafana stack, per-component enable/disable, lightweight defaults, documented min resources	internal/localnet/observability.go, internal/localnet/observability_overlay.go, assets/compose/, assets/grafana/; docs: docs/observability.md
Canton-specific Grafana dashboard presets (TPS, command completion latency, active contract counts, per-template throughput)	assets/grafana/, assets/dashboard_test.go; docs: docs/dashboard-customization.md
dpm localnet metrics subcommand (Grafana URLs + text summary: throughput, latency p50/p99, resource usage)	internal/cli/localnet/metrics.go, internal/metricsq/queries.go
DAR management CLI (upload/list/info/download/diff/remove/build-upload/watch), multi-participant, dpm build integration, SCU-aware diff	internal/cli/localnet/dar/, internal/localnet/darops/, internal/dar/ (incl. diff.go, lf_inspect.go)
DAR Web UI (drag-and-drop upload, per-participant vetting toggles, package explorer tree, diff viewer, hot-deploy indicator)	frontend/src/screens/DARScreen.tsx, DARPackageTree.tsx, DARDiff.tsx
Contract tracking CLI (contracts watch, tx ls/replay) on Ledger API v2 (live watch, multi-filter tx listing, per-party visibility projection)	internal/cli/localnet/contracts.go, internal/cli/localnet/ledger_filter.go, internal/canton/ledger/ (stream.go, filter.go, project.go)
Contract tracking Web UI "Explorer" (live ACS table, transaction timeline, contract detail drawer, per-party visibility projection)	frontend/src/screens/ExplorerScreen.tsx, ContractDetailDrawer.tsx, TxReplayDrawer.tsx; docs: docs/explorer.md
Optional AI agent skill documents (safe dpm localnet workflows: lifecycle, DAR upload, package inspection, contract queries, log/status)	internal/skills/docs/ (localnet-lifecycle.md, dar-upload.md, inspect-contracts.md, hot-deploy.md, ci-localnet.md), internal/skills/skills.go;
Documentation (usage, dashboard customization, DAR workflows, contract explorer)	https://bitdynamics-ab.github.io/canton-devkit/

Observability Support Evidence

built-in UI

Grafana Dashboard

Adoption

At least 5 companies/teams have started using it in their daily Canton development workflow:

• Gateway.FM
• NODESJUMPER
• FTP
• ITRockets
• Moonsong Labs

Notes for reviewers

Live demo available on request. Ready for milestone review whenever the committee is. Thanks!

[hythloda]

/vote

[git-vote]

Vote created

@hythloda has called for a vote on Canton devkit (dpm component) - a unified toolkit for Canton app development (LocalNet, dar helpers, etc) #18 Milestone 2: Web UI, Observability, Monitoring, DAR & Contract Tooling, Optional AI Agent Skill Documents (#387).

The members of the following teams have binding votes:

Team
@canton-foundation/tech-ops-voting

Non-binding votes are also appreciated as a sign of support!

How to vote

You can cast your vote by reacting to this comment. The following reactions are supported:

In favor	Against	Abstain
👍	👎	👀

Please note that voting for multiple options is not allowed and those votes won't be counted.

The vote will be open for 21days. It will pass if at least 51% of the users with binding votes vote In favor 👍. Once it's closed, results will be published here as a new comment.

[Denend]

We’ve been using canton-devkit as part of our Canton development workflow. LocalNet lifecycle management (up/down/status/logs) and environment export for app configuration. The machine readable status output and enriched doctor diagnostics made headless automation straightforward

[sibuser]

I can confirm that Gateway.FM team has been using devkit to run e2e test canton apps with LocalNet instances.

[dmitryk-cf]

Our team has also been using canton-devkit as part of our regular Canton development workflow. LocalNet startup, environment export, and the status/doctor commands have made it easier to run and debug our app/test setups without extra manual steps.

[Jatinp26]

@srikanth-bitdynamics Do you have rest of 2 teams, as per adoption metric of 5 in this milestone, using devkit?

[srikanth-bitdynamics]

@Jatinp26 yes I've asked them to drop a note here so they should comment here by Wednesday's call hopefully

[itrocket-team]

Tried canton-devkit - a single localnet up brought up a full local Canton network and worked exactly as documented. It looks useful for our Canton-related work, and we plan to use it in our development process.