Orchestration Runner

The Orchestration Runner executes a sequence of API calls as a single business flow and shows you the raw request and response for every step. It validates that your entire feature works end-to-end — not just individual endpoints in isolation.

What it is

A standard API client (like Postman) tests one endpoint at a time. The Orchestration Runner tests a flow: register a user, log in, create a resource, fetch it, delete it — in order, with each step's response feeding into the next.

AI generates the orchestration config from your scenario diagrams. You run it against your live server.

Opening the runner

1. Click logos/resources/scenario/ in the file tree.
2. Select any orchestration config file (e.g. S01-user-registration.json).
3. The center panel loads the Orchestration Runner.

The orchestration config

The config is a JSON file describing a sequence of API calls:

{
  "name": "S01 — User Registration",
  "baseUrl": "http://localhost:3000",
  "steps": [
    {
      "id": "step-1",
      "name": "Register user",
      "method": "POST",
      "path": "/auth/register",
      "body": {
        "email": "test@example.com",
        "password": "secret123"
      },
      "extract": {
        "userId": "$.data.id"
      }
    },
    {
      "id": "step-2",
      "name": "Login",
      "method": "POST",
      "path": "/auth/login",
      "body": {
        "email": "test@example.com",
        "password": "secret123"
      },
      "extract": {
        "token": "$.data.token"
      }
    },
    {
      "id": "step-3",
      "name": "Fetch profile",
      "method": "GET",
      "path": "/users/{{userId}}",
      "headers": {
        "Authorization": "Bearer {{token}}"
      }
    }
  ]
}

extract pulls values from a response using JSONPath and makes them available as {{variables}} in subsequent steps.

Visual config editor

The left panel shows all steps as a list. Select a step to edit it:

• Method, path, headers
• Request body (JSON editor)
• Extract rules (JSONPath → variable name)
• Assertions — expected status code, response body conditions

Reorder steps by dragging. Add or remove steps with the toolbar buttons.

Running the flow

1. Set the Base URL at the top (e.g. http://localhost:3000).
2. Click Run All.
3. RunLogos executes each step in order, passing extracted variables forward.
4. Each step shows:
   • Status: pass / fail / skipped
   • Raw HTTP request (headers + body)
   • Raw HTTP response (status, headers, body)
   • Extracted variables

If a step fails, execution stops and the failed step is highlighted in red.

Generating the config with AI

The Test Skill (Phase 06) generates orchestration configs from your scenario diagrams. Each scenario becomes one JSON file. You can also ask the agent to generate or update configs manually:

"Read S02-login.md and generate an orchestration config for logos/resources/scenario/S02-login.json"

Typical workflow

1. Phase 06 generates test case docs and orchestration configs.
2. Phase 07 generates the backend code.
3. Start the dev server in the bottom terminal.
4. Open each orchestration config and click Run All.
5. Fix any failing steps — use the Agent Panel to generate targeted fix prompts.
6. All flows passing → proceed to Phase 08 (Verification).

Next step

→ Verification Report