> ## Documentation Index > Fetch the complete documentation index at: https://docs.ironbee.ai/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > IronBee (ironbee.ai) is an AI QA engineer: it verifies code changes against the running app and keeps the evidence. It is not related to the IronBee open-source web application firewall. > The console is at https://console.ironbee.ai. > Install the CLI with `npm install -g @ironbee-ai/cli`, sign in with `ironbee login`, and set up a project with `ironbee install`. > In GitHub workflows, pin the action to `ironbee-ai/ironbee-action@v1.1.0` and pass the `IRONBEE_API_KEY` secret as the `ironbee_api_key` input. > Console scenarios (https://docs.ironbee.ai/console/scenarios) and CLI saved scenarios (https://docs.ironbee.ai/cli/guides/scenarios) are separate features. > For plans and prices, link to https://ironbee.ai/pricing; the website is the source of truth for them. # Write a scenario > Build a scenario in the editor from sections and rows, set its inputs and outputs, and save it. You write a [scenario](/console/scenarios) in the editor, as rows of plain-language instructions. Open it with **Edit** on a scenario, or by creating one with **New scenario** or **Duplicate**. Scenario editor with the name, Discard, Save changes and Run scenario at the top, the Show last run switch and the action and node counters, the Outline on the left, and the Setup, Main and Teardown sections with Import, Assert, Wait and Act rows Scenario editor with the name, Discard, Save changes and Run scenario at the top, the Show last run switch and the action and node counters, the Outline on the left, and the Setup, Main and Teardown sections with Import, Assert, Wait and Act rows The editor has: * the scenario's name, which you can edit in place, and its type * **Discard**, **Save changes** and **Run scenario**, with the save status next to them: **Unsaved changes**, or when and by whom the scenario was last saved * a toolbar with **Show last run**, undo and redo, the action and node counters, **Scenario settings** and the keyboard shortcuts * the **Outline** on the left, to jump to a section or a row * the **Setup**, **Main** and **Teardown** sections. See [How a scenario is built](/console/scenarios#how-a-scenario-is-built) *** ## Kinds of rows | Kind | What it does | Settings | | - | - | - | | **Act** | Does something in the app, like clicking, typing or opening a page | **If it is blocked**: **Don't retry**, or retry up to 3 times. Retry only actions that are safe to repeat | | **Assert** | Checks that something is true. The agent is told not to change the app | **If it fails**: **Blocking** fails the run and skips the rest of **Main**. **Soft** records the failure and the run goes on | | **Extract** | Reads a value from the page and saves it under a name for the actions after it | **Save as**: one or more names, separated with commas | | **Wait** | Looks again at intervals until something shows up or changes. Running out of time fails it | **Give up after**, from 60 s (the default) to 30 min, and **Check every**, from 5 s to 5 min (10 s by default) | | **Step** | Groups rows under a name. It runs nothing itself | **Description**, shown under the Step's name | | **Import** | Runs another scenario's **Main** here with its inputs, and uses what it gives back | See [Imports](/console/scenario-references#imports) | Write each **Instruction** the way you'd tell a teammate, in up to 4,000 characters. Type `@` to use an input, a saved value, an output of an import, a variable, a secret or a property in it. See [References](/console/scenario-references#references). Every row has a **Name**, shown in the outline and in run results. It follows the instruction until you change it, and it's unique inside its section or Step. An Act does something; an Assert checks that something is true and fails the run when it isn't. When an Act reads like a check, the editor offers **Make it an Assert**. *** ## Add and arrange rows To add a row, click **Add** at the end of a section, **Add inside** under a Step, or **+** before a row. The **Add a row** menu lists: * **Actions**: **Act**, **Assert**, **Extract** and **Wait** * **Structure**: **Step** and **Import** * **Quick starts**: **Sign in with a secret**, **No error is shown** and **Page has loaded** Pick one with the arrow keys and Enter, or press 1–6. The row opens in a dialog: fill it in and click **Add**, or **Cancel**. Add a row menu with Act, Assert, Extract and Wait under Actions, Step and Import under Structure, three quick starts, and a preview of what the selected kind adds Add a row menu with Act, Assert, Extract and Wait under Actions, Step and Import under Structure, three quick starts, and a preview of what the selected kind adds Click a row to edit it in the same dialog, and click **Done** to keep your changes. The **...** menu on a row offers **Edit**, **Duplicate** and **Delete**. Drag a row to move it, into a Step or out of one. With the keyboard, move a row with Alt and the arrow keys. See [Keyboard shortcuts](#keyboard-shortcuts). *** ## Scenario settings Click **Scenario settings**, at the top of the outline or in the toolbar: | Setting | What it's for | | - | - | | **Description** | One line shown under the name. The scenarios list searches it | | **Tags** | Comma-separated, to filter the scenarios list. Up to 16 tags, each up to 64 characters | | **Inputs** | Values a run passes in. An input without a default is required | | **Outputs** | Values this scenario gives back to the scenarios that import it | Scenario settings dialog with a description, tags, an optional input with its default value, and the Outputs section Scenario settings dialog with a description, tags, an optional input with its default value, and the Outputs section Inputs and outputs are covered in [Inputs, references and imports](/console/scenario-references). *** ## Problems and warnings The editor checks the scenario as you type, the same way a save does. What it finds shows on the rows, in the outline and above the sections. * **Errors** break the scenario's rules, and IronBee refuses to save them. For example: **Main** has no action, a Step is empty, the scenario has more than 60 actions, or two scenarios would import each other. * **Warnings** don't stop saving, but a run that reaches the row ends in a run error, or stops before it starts. For example: * a row reads a value before the row that writes it * a secret, variable or property isn't defined in any environment, or only in some * a reference can mean two secrets *** ## Save your changes Click **Save changes**, or press ⌘S (Ctrl+S). **Discard** drops the changes you haven't saved. If you clicked it by mistake, undo it. * If someone else saved the scenario after you opened it, the editor lists **Their changes** and asks what to do: * **Load latest** replaces your edits with their version. * **Save mine anyway** replaces their changes with yours. * **Keep editing** closes the dialog. * If you try to leave with unsaved changes, a **Leave without saving?** dialog asks you to confirm. * **Run scenario** saves your changes first, then opens the [run dialog](/console/scenario-runs#start-a-run). A saved change applies from the next run. Scenarios that import this one also run the new version from their next run. A save that would break one of them is refused, and the editor names the scenario it would break. *** ## Keyboard shortcuts Click the keyboard icon in the toolbar to see them in the editor. On Windows and Linux, use Ctrl for ⌘ and Alt for ⌥. | Action | Keys | | - | - | | Insert a reference | `@` | | Done in a dialog | ⌘ Enter | | Move up or down | ⌥ ↑ / ⌥ ↓ | | Move into the Step above | ⌥ → | | Move out of the Step | ⌥ ← | | Duplicate | ⌘ D | | Undo | ⌘ Z | | Redo | ⇧ ⌘ Z | | Save | ⌘ S | | Close a menu or a dialog | Esc | *** ## Limits | What | Limit | | - | - | | Scenarios in a project | 100 | | Actions in a scenario, counting what its imports bring | 60 | | Nodes in a scenario: every row, counting Steps, Imports and what imports bring | 100 | | Levels of Steps and Imports | 8 | | Scenario name | 255 characters | | Description | 1,024 characters, on one line | | Instruction | 4,000 characters | | Row, input, output and import names | 64 characters | | Saved values and import outputs in a scenario | 50 | | Retries on an Act | 3 | The counters in the toolbar show how many actions and nodes the scenario uses. A counter appears under a field as it nears its limit. *** ## What's next? Pass values in and out, and reuse one scenario in another. Pick an environment, pass inputs and read the run. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.