Skip to content

Proposal: Canton Multi-Party Workflow Choreography Reference Implementation

CLOSEDPull Request
by blackthornlover21-03-2026Declined
2M CC requested

Development Fund Proposal Submission

Proposal file: /proposals/canton-multiparty-workflow-choreography.md

---

Summary

Canton Multi-Party Workflow Choreography (CMWC) delivers a reusable, Canton-native reference implementation for coordinating workflows across parties who cannot see each other's state. The project ships four Daml choreography primitives (WorkflowHandoff, LocalAttestation, AwaitingStep, WorkflowTerminal), a per-participant TypeScript workflow runner, a fully worked four-party equity settlement reference scenario, and a choreography design guide — across 4 adoption-based milestones requesting 2,000,000 CC in total funding.

---

Checklist

  • [x] Proposal file added under /proposals/
  • [x] Milestones and funding amounts defined
  • [x] The proposal includes adoption-based milestones with verifiable acceptance criteria
  • [x] Acceptance criteria included
  • [x] Alignment with Canton priorities described

---

Notes for Reviewers

  • This proposal addresses the coordination problem created by Canton's own privacy model: when parties cannot see each other's state, how does a multi-party workflow advance correctly? This problem does not exist on public chains and cannot be solved by importing patterns from other ecosystems.
  • The reference scenario — four-party equity settlement with CCP novation — is chosen because it contains every choreography design pattern (party-set transitions, off-contract attestations, parallel instruction issuance, terminal-state finality) in a single commercially recognizable workflow.
  • The WorkflowHandoff primitive enforces at the Daml model level that no prior-step party appears on a new-step contract unless explicitly listed — the most common correctness failure in Canton multi-party workflows today.
  • This proposal is a natural complement to CCDM (PR #117): CCDM handles cross-synchronizer messaging; this proposal handles cross-privacy-domain choreography within a single synchronizer.
  • The four-party settlement flow is included as the proving scenario for the choreography patterns, not as the product itself. The funded deliverable is the reusable open-source implementation package around that scenario: the Daml pattern library, participant runner skeleton, verification suite, and adaptation guide.