7.4The steps you can use

obsrvly runs the steps from top to bottom. It stops at the first step that fails. A selector is a short text that points to one element on the page, such as a button (see 7.5).

Step	Fields	What it does
`goto`	`url`	Opens the page and waits until it has finished loading. Only `http://` and `https://` addresses work.
`click`	`selector`	Waits for the element, then clicks it.
`fill`	`selector`, `value`	Types the value into a field. It replaces what is already there. The value can be a team secret (see 7.8).
`wait_for`	`selector`	Waits until the element is on the page and visible.
`assert_visible`	`selector`	Checks that the element is visible. It waits for it, and fails if it does not appear in time.
`assert_text`	`selector`, `contains`	Checks the text of an element. It reads the text of the first matching element. Passes if it contains the expected text. Capital letters must match.

In the dashboard, pick the step from the list on each row. The list shows the names in the first column, such as wait_for and assert_text. Each row then asks for its fields:

goto: url.
click, wait_for, assert_visible: selector.
fill: selector and a value. If your team has secrets, a secret… list appears next to the value.
assert_text: selector and the expected text.

Through the API, the steps are a JSON list. Each step has a kind and its fields, as in 7.6. Every step except goto can also take timeout_ms (see 7.7).

When a step fails#

These messages name the step, counting from 0 (see 7.9).

Message	What it means
`text "Welcome back" does not contain "Dashboard"`	`assert_text` found the element, but its text is different. The message shows up to 80 characters of what it found.
The step's error says it timed out	The element did not appear in time. Check the selector (7.5) and the timeouts (7.7).
`page answered 500`	A page the script opened answered with an error status. Check your site.
`status 503`	The first page opened answered with an error status (400 or higher). Check your site.
`after the last step: page answered 500`	The page answered with an error after the last step ended.
`only http and https URLs are allowed, got ftp:`	A `goto` step uses another kind of address.
`not a valid URL: <url>`	A `goto` step has an address that is not complete.

If the script itself has a mistake, the check fails before any step runs. The message starts with script parse:. For example:

script parse: step 0 (goto): missing url
script parse: step 1 (click): missing selector
script parse: step 2 (fill): selector and value required
script parse: step 3 (assert_text): selector and contains required
script parse: step 4: unknown kind "hover"
script parse: script must be a JSON array

Page checks after the last step (API only)#

Through the API, you can add page checks in assertions. obsrvly runs them after the last step. Browser monitors accept only these four types (see chapter 4):

Type	What it checks	Message when it fails
`status_code`	The status of the first page opened. Use eq, neq, gt or lt (as in chapter 4).	`assertion status_code == 200: got 503`
`body`	The visible text of the last page. Use eq, neq, contains or matches (a regular expression).	`assertion body contains "Welcome" failed`
`response_time_ms`	The time the whole script took. Use eq, neq, gt or lt.	`assertion response_time_ms < 5000: got 6120`
`no_console_errors`	Fails if the page logged any error in the browser console. It needs no `op` or `value`.	`2 console error(s): <the first error>`

Any other type is refused when you save, for example assertions[0]: type "header" is not one of status_code, body, response_time_ms, no_console_errors.

If you add a status_code check, it decides whether the first page's status is good. Without one, a first page that answers with 400 or higher makes the check down.