Scenarios
This is the app’s landing screen: every scenario in the current workspace as one list. Scenarios are created, searched, opened in the editor and deleted here.
Scenarios
Scenarios
12 scenarios · 3 active · 48 runs today · 92% success
| Name | Trigger | Status | Runs 7d | Success |
|---|---|---|---|---|
| Telegram support | Telegram · @acme_bot | Active | 128 | 94% |
| Nightly report | Schedule · 0 3 * * * | Active | 7 | 100% |
| Inbox triage | Manual | Draft | — | — |
The summary line
Four numbers sit under the heading: how many scenarios there are, how many are active, how many runs happened today, and what share of runs succeeded.
The first two follow the current search. The last two describe the whole workspace and ignore the filters: “runs today” counts from the start of the current day (UTC), and “success” is the share of runs that completed successfully over the last seven days.
Tabs and search
The All, Active, Paused and Draft tabs filter the list. The counter next to a tab is computed over the current search and does not depend on which tab is open — that is what makes all four numbers meaningful at once.
A scenario counts as active when it has a deployed version; everything else is a draft. The server does not distinguish a separate “paused” state today, so that tab is always empty and its counter is zero.
Search runs on the server over the name and the description, case-insensitively. The request is sent after a short pause following your last keystroke. Changing the query or the tab returns the list to the first page.
The list is sorted by creation time, newest first, and paged 20 cards at a time; “Prev”, “Next” and a range like “1–20 of 34” sit at the bottom.
What a card shows
- Status dot and badge — “Active” (a deployed version exists) or “Draft”.
- Name, and under it the trigger derived from the deploy configuration: “Schedule · <cron>”, “Telegram · @bot”, “Webhook”, “Chat” (public chat or the widget), or “Manual” when nothing is deployed.
- Runs 7d — runs over the last seven days; Success — the share of successful ones among them. With no runs at all both read “—”.
- Version number (
v3) — the latest saved version of the scenario, which is not necessarily the deployed one. - Trash icon — appears on hover and opens the delete confirmation.
Clicking anywhere on the card except the trash icon opens the graph editor.
Creating a scenario
The New Scenario button opens a dialog with a single field.
New Scenario
New Scenario
or start from a template:
| Template | Summary |
|---|---|
| Простой LLM-чат | Entry, an LLM agent and a reply. The basis of any chatbot. |
| RAG по базе знаний | Search the workspace documents and answer from them. |
| Агент с веб-поиском | Answers questions that need fresh data from the internet. |
Type a name and press Create — the editor opens with an empty graph.
Below the field is the “or start from a template” section. Clicking a template creates a new scenario with a graph already assembled; when the name field is empty, the template’s own title is used. The result is an ordinary scenario: you can edit, save and deploy it, and it no longer refers to the template.
Deleting
The trash icon on a card opens the “Delete scenario?” dialog. Confirming removes the scenario together with all of its versions and its execution history.