José Miguel Gómez

Back to the home page

Kata hexagonal + DDD

One game-jam core on REST, GraphQL and PGlite, measured against layers

The code lives in a private repository; this sheet and the demo, when there is one, show the work.

Status
Coming soon
Type
Developer tool
Running since
November 2026
Role
Architect and sole engineer
Domains
software architecture, domain driven design, API design
Cast
TypeScript, Nodejs, Fastify, Graphql, Kysely, PostgreSQL, Pglite, Vitest, Hurl

Context and problem

Hexagonal and DDD are argued about without numbers. This kata builds the same KataJam contract twice, hexagonal + DDD and a layered control, and measures what five business changes cost each one.

How it was built

Built in 4.1 h · USD 85.34 of AI · 425 tests · 4 audits

Source: Trysac/l13-kata-hexagonal-ddd docs/development-report.md (lines 12, 15, 22), dated 21 November 2026.

Results with dated evidence

Outcomes as stated in the manifest; the measured figures are the ones with a date.

  • Both implementations pass the same 59 black-box Hurl files (46 base scenarios plus 13 change-scenario files) with 473 valid CloudEvents (make ci).
  • Five change scenarios measured on both: the hexagonal side touched 40 files and broke 4 tests, the control 20 files and 0 tests (metrics/cs).
  • Zero dependency-rule violations under hexagonal and Clean rule sets; domain line coverage 98 %.

Case study in preparation

The full case study adds these sections when it is published:

  • Constraints
  • Architecture diagrams (C4 context and containers)
  • Three key decisions
  • What went wrong
  • Demo video
  • What I would do differently