How to test a scenario in Scenario Builder?

Scenario Builder is a new module that replaces Scenario Dialog. This guide helps you run a scenario to verify it works, view run history, and use the chat preview.

When you need it

• You need to verify the scenario works correctly after changes.
• You want to see how the dialog looks for the bot user.
• You need to find an error in the logic or node configuration.

What you need to know

• Full run — the scenario runs from the beginning (Start Node). During execution you can interact via the chat preview.
• Run history (Runs) — the left panel stores history of all runs. For each run you can see executed nodes and status.
• Chat preview — a widget that shows the dialog in real time during the test. You can send messages and see bot responses.
• Test single action — for Action nodes you can test the action without running the full scenario (see Use action in scenario).

Before you start

• You are logged in with integrator or administrator rights.
• You have an open scenario in Scenario Builder.
• The scenario is published (no critical validation errors).

Step-by-step instructions

1. Run the scenario

1. In the top toolbar, click «Run» or «Test».
2. The system validates the scenario. If there are errors — the run is blocked and an error list is shown.
3. After successful validation, a new run is created. The execution indicator is displayed (running, success, error).

2. Interact via chat preview

1. During execution, open the chat preview (if it does not open automatically).
2. The chat widget shows bot messages and lets you send a response (text, buttons).
3. Send a message — the scenario continues execution according to its logic.

3. View run history

1. In the left panel, find the «Runs» section.
2. Select a run from the list — information about executed nodes and status is displayed.
3. You can view execution logs if the run ended with an error.

4. Find a block in the execution sequence

1. Open a run in the «Runs» section — the «Execution Details» panel with the sequence of executed nodes appears.
2. In the search field above the sequence, type the block name or its ID. To find a block by exact ID, type # followed by the ID, for example #12.
3. To search only blocks of certain types, click the funnel button to the right of the field and check one or more types, for example «Router». The list contains only the types of blocks that ran in this run, with the number of steps. With an empty field, every step of those types is a match. The chosen types appear under the field: the cross removes a type, «All types» resets the choice. The crossed-out eye button next to the arrows keeps only the found steps in the list.
4. Press Enter to go to the next match or Shift+Enter to go to the previous one. The search covers only the nodes that ran in this run. If a node ran several times, each pass is a separate match, and the panel shows, for example, «Pass 2 of 3».
5. The «List» and «Canvas» buttons (icons on the right) define where to show the found node: scroll the list to it, open it on the canvas, or both. You cannot switch both off.
6. The «Deleted» tag means the node ran in this run but is no longer in the scenario, so it can't be shown on the canvas. In the type list, such steps are grouped under «Deleted blocks».
7. Esc clears the search text; the chosen type stays.

5. View the input and output data of a step

1. In the «Execution Details» panel, click a step in the sequence or open it in a separate window with the button in the step row.
2. The «Input data» block shows what the step worked with: the action name, the step parameters as configured in the block, the variables used in parameters with their values right before the step, and the rest of the variables before the step.
3. The «Output data» block shows what the step passed on: the step result, the chosen exit — the block the scenario actually went to (for example, «Exit: default → block #8») — and the changed variables tagged «added», «changed» or «removed». If the step failed, its error text is shown here. If the step result is error but there is no error text, look for it in the «Logs» of this step. A step that waits for the user's reply (buttons, input, list search) shows the «Waiting» mark instead of an exit: where the scenario went after the reply is visible in the next step.
4. Run service keys (channel, chat_id, section_id, test_run_id, initial_state) are hidden; the «Hidden service keys» line shows how many.
5. To see the full JSON — «State Before», «Result», «State After» — switch on «Show raw payload» above the blocks. The choice is remembered in the browser. «Output Messages» and «Logs» are visible in both modes.

Important. «Input data» is the parameters as configured plus, separately, the values of the variables they refer to. It is not the string after variable substitution: if a parameter says Discount {{discount}}%, you see exactly Discount {{discount}}% and the value of discount next to it (for example, 10), not the final Discount 10%. Variables that are not in the dialog state (for example, bot constants) are marked «not in state».

6. Stop execution (optional)

1. During execution, click «Stop» in the toolbar.
2. Execution is interrupted; the state is saved in run history.

7. Publish changes after testing

1. If you found changes needed during the test — make them in the scenario.
2. Click «Publish» in the toolbar. In the «Publish Scenario» window, optionally add a «Version Comment» and click «Publish».
3. Run the test again to verify.

Related materials

• View version history
• Use action in scenario — test a single action in Node Inspector
• Find a node on the canvas — search across all scenario nodes, not only the executed ones
• Scenario Builder hub