Platform Evolution
Assessment run-2026-07-27-1 · 213 of 213 repos assessed
Repos / tc-document-platform

tc-document-platform

.NET 10 document-generation platform (quotes, invoices, certificates, commission statements) built as MassTransit-over-SQS services with Liquid templates in MongoDB and PDFreactor rendering, deployed as containers on AWS ECS.

96 · greenlowloosely attachedAWSZDD: no excuseservice

Health score

96

0 to 100 across seven weighted factors; nothing is hidden. Green is 75 and above, amber 50 to 74, red below 50. How scoring works.


Runtime currency / 25
25
Runs on supported modern .NET
Legacy transport / 15
15
No legacy WCF or SOAP surface
Test posture / 15
15
Good test coverage
Security hygiene / 15
15
No credentials found in source
Dangerous coupling / 15
11
depends on pinning component(s): application-framework
Activity / 10
10
Actively maintained
Deployment hygiene / 5
5
CI in place
Health over time
2026-07-27 (run-2026-07-27-1): 96
Baseline 96 recorded 2026-07-27. The trend fills in with each monthly assessment run.

Facts

Languagescsharp
FrameworksASP.NET Core 10, MassTransit 8, FastEndpoints 8, Carter, Fluid/Liquid, Aspire, AWS Lambda
Runtimes.NET 10
Tests / CIgood / CI in place
Last commit2026-07-24
OwnerOwner unknown (mapping to be curated)
DomainBookings and fulfilment (system generated; curate in repo-owners.json to override)
Product areaShared platform
HostedAWS Deployed to AWS
  • build "tc-document-platform" (azure-pipeline.yml): AWS signal "ecsTerraformRepo"
  • build "tc-document-platform-lambdas" (azure-pipeline-lambdas.yml): AWS signal "ecsTerraformRepo"
  • build "tc-document-platform-pdfreactorcore" (azure-pipeline-pdfreactorcore.yml): AWS signal "ecsTerraformRepo"
  • build "tc-document-platform-management-api" (azure-pipeline-management-api.yml): AWS signal "ecsTerraformRepo"
Releases0 release pipeline(s):
Integrationrest, masstransit, sqs, jwt
SourceOpen in Azure DevOps

Risk flags

No risk flags recorded.

Each flag links to its glossary definition.

Dependencies in context

Laying out the graph…
greenamberreddata store (square, dashed link)Node size follows fan-in. Click any node to open it.

Entanglement

loosely attached Bound into the core by 1 mechanism(s).

Detaches when its pinning dependencies are cut or multi-targeted (application-framework).

Migration playbooks

Generated from this repo's assessment data: what is feasible, what blocks it, and the steps in order. Ratings: green means actionable now, amber means sequence other work first, red means blocked, grey means not applicable.

ZDD adoption

No excuse

Nothing technical is holding this repo on the three-week release. It either already deploys without downtime (AWS) or is a modern, well-tested service the automated stealth path can carry today. Already deploys as a container to AWS, which releases without downtime today. Separating it from the three-week release is a process decision, not a technical one.

Dependencies and gotchas
  • Serialized state or messages in flight (Redis or queues): new code must deserialize objects written by the previous version
Steps
  1. Adopt the four-stage discipline from the ZDD engineering standard: pre-deployment compatibility, deployment, feature activation behind a flag, post-release cleanup.
  2. Make every change dual-version safe: backwards-compatible APIs (new major version for breaking wire changes), expand and contract for any schema change, and consumers tolerant of unknown fields.
  3. Audit serialized data: anything in Redis, queues or shared caches must deserialize cleanly across adjacent versions.
  4. Wrap behaviour changes in feature flags so code deploys dark and release becomes a business decision, not a deployment event.
  5. Use ECS rolling deployments with health checks gating traffic, and take this service off the three-week release train.
  6. Verify with area-scoped regression rather than the whole-platform cycle, and watch error rates after rollout with automated rollback on alert.

SQL Server exit

Not applicable

No SQL Server dependency is recorded for this repo. Nothing to migrate.

Containerisation

Already containerised

Its pipelines already build containers and deploy to AWS. Nothing to do here.

Recommended actions

No specific actions for this repo. It either needs only normal maintenance, or its work is covered by an estate-level action. See the full programme.

Assessment notes

tc-document-platform

Purpose

The estate's next-generation document platform: callers submit document definitions via TC.DocumentationBuilder.Client/Facade, MassTransit choreographs TemplateStore, DefinitionStore, Engine and Generator services over Amazon SQS/SNS (with an S3 claim-check for large payloads), Liquid templates stored as MongoDB data are rendered by Fluid, and PDFreactor produces the final PDFs. Document families include Quote, Invoice, Best Ever Month Certificate, TC Monthly Commissions and TC CoPilot documents, plus a Management Portal API serving the phenix-ui document frontend and an audit store.

Tech stack

C# on .NET 10 / C# 14 with central package management: MassTransit 8.5 (SQS/S3/Mongo), FastEndpoints 8 and Carter, Fluid.Core templating, i18Next localisation, MongoDB.Driver 3.8 via ConnectionBroker, TC.Cognito.Auth.Client bearer auth, Microsoft.Extensions resilience/hybrid caching, an Aspire AppHost (Mongo/Postgres/LocalStack) for local orchestration, and a MapVisuals pushpin-rendering AWS Lambda under lambdas/. A commercial PDFreactor engine runs in its own Docker host image (tc-docplat-pdfreactorcore).

Deployment

Included via build->yaml-deploy with 0 classic releases and four build pipelines: tc-document-platform (ten containerised apps — TemplateStore, Engine, Facade, Generator.PdfReactor, DefinitionStore, BemcBuilder, TCCopilot, Quote, Invoice, TCMonthlyCommissions — plus five NuGet packages), tc-document-platform-lambdas, tc-document-platform-management-api and tc-document-platform-pdfreactorcore. All use the shared InternalOSS templates, cx container namespace, ECS Terraform from aws-cx-tc-apps (eu-west-1, functional area cx).

Dependencies

Internal: TC.ConnectionBroker.* (tc-connectionbroker), TC.Diagnostics.HealthChecks (tc-diagnostics-portal), TC.Phenix.ApplicationFramework (application-framework), TC.Destination.Queries.API.Client and TC.DestinationManagement.Queries.GraphQL.Client (phenix-destination-management), TC.Public.Assets.Api.Client (tc-public-assets-api), TC.CommissionStatements.DataContract (commission-statements-api). External: PDFreactor (licensed renderer), Trustpilot (API key pulled from SSM at /docplat/trustpilot/trustpilot_api_key), AWS Cognito for bearer auth.

Data stores

MongoDB via ConnectionBroker targets: DocumentDashboard (audit), plus template and definition stores; MassTransit also uses Mongo for sagas/outbox. S3 document-definition-claim-check-bucket holds large message payloads with a 24-hour TTL. Generated document assets are stored and served by external id/correlation id through the Management Portal API.

Integration points

Inbound REST (Facade for generation/status — its HTTP+JSON contract with TC.DocumentationBuilder.Client is explicitly guarded; Management Portal API bearer-protected for phenix-ui), asynchronous MassTransit messaging over Amazon SQS/SNS between the platform's own services, and published client/message NuGet packages for consumers.

Code health

Excellent: ~147 test files across unit and integration test projects (integration tests run in CI with a generated dev cert), an Aspire AppHost for realistic local runs, and unusually thorough in-repo engineering documentation (CLAUDE.md mirrored to .cursor/rules, covering client compatibility, template-as-data philosophy and UI integration). Very active — last commit 2026-07-24.

Modernisation risks

Low — this is the modern target other documentation systems (e.g. the legacy backoffice-documentation-* family) should converge on. Watchpoints rather than risks: a complex four-pipeline, ten-app deployment surface; the licensed PDFreactor dependency in its own container; strict client/Facade contract compatibility constraints; and template partials embedded in assemblies that bypass the template-editor data model.