Test Management

The platform gives every ruleset its own test suite — same YAML format as ordo-cli, with a single source of truth across Studio, CLI, and CI.

Case Structure

ruleset: discount-check
cases:
  - name: vip user gets 20% discount
    input:
      user: { id: u1, vip: true, age: 28 }
      order: { amount: 200 }
    expect:
      code: VIP
      output: { discount: 0.2 }

  - name: minors are denied
    input:
      user: { id: u2, vip: false, age: 16 }
      order: { amount: 50 }
    expect:
      code: DENY
      output: { reason: 'underage' }

API

Operation	Endpoint
List	GET /api/v1/projects/:pid/rulesets/:name/tests
Create/update	POST/PUT /api/v1/projects/:pid/rulesets/:name/tests[/:tid]
Run one	POST /api/v1/projects/:pid/rulesets/:name/tests/:tid/run
Run all	POST /api/v1/projects/:pid/rulesets/:name/tests/run
Project-wide run	POST /api/v1/projects/:pid/tests/run
Export YAML	GET /api/v1/projects/:pid/rulesets/:name/tests/export

Coupling with Releases

When a release request is created, the platform automatically runs all test cases for the affected rulesets. Any failure blocks creation of the release request.

You can disable auto_run_tests in a release policy to skip this gate, but production usually shouldn't.

CI Integration

• Export YAML from the platform and commit it to your code repository.
• Run via ordo-cli during PR checks:

ordo test --rules ./rulesets --tests ./tests --reporter junit > junit.xml

Output formats: JUnit XML, JSON, TAP — all directly consumable by GitHub Actions / GitLab CI.

Trace & Failure Diagnosis

When a test fails, the platform returns the full execution trace. Click the result in Studio to see:

• Expected vs actual output code
• The last branch that matched before the divergence
• Each action node's assignment trail

See Studio Editor — Trace Panel.