Automatic UI actions are sourced from a XML file, and run sequentially once the UI has been fully built.
These actions are designed for machine usage and macros. Typical scenarios can be: - To replace human input in automatic testing pipelines.
- In documentation, to show users how to do something within the application. - To allow LLM agents or crawlers to operate on a UI built with vs.

Supported actions#

• action.sleep wait for some time before proceeding.
• action.click to run the callback associated to a widget.
• action.fill to add some value as keyboard input or copy/cut.
• action.tab
• action.focus
• action.wait to wait for some condition on the ui to be true.

Special blocks for testing#

• test the test is resolved.
  • deny the test ends and is marked as unresolved.
  • accept if unresolved, continue with the next operation.
  • exit to end the application.

Support in documentation#

Documentation for vs projects is intended to be written in Markdown.
As such, XML code can be directly embedded in it, which means vs components are usable (please notice that based on the local policies, this functionality might be disabled).
The special component fl:inline-helper can embed UI actions in it, which will run on request.

Additional props to handle semantics#

Additional explicit information should be provided to ensure the semantic of the exposed UI components is explained.
This is done by nesting a semantics element to each element needing that. Elements with semantics specified should also have a name attribute.
For example:

<fl:container>
<script>
    void on_new(){}
    void on_open(){}
    void on_delete(){}

    $cb(on_new)
    $cb(on_open)
    $cb(on_delete)
</script>
<semantics>
    <callback name="on_new">Explanation</callback>
    <callback name="on_open">Explanation</callback>
    <callback name="on_delete">Explanation</callback>
</semantics>
<fl:button on.callback="on_new" label="New"/>
<fl:button on.callback="on_open" label="New"/>
<fl:button on.callback="on_delete" label="New"/>
</fl:container>

This additional context is extremely good to automatically generate documentation, but also to ensure proper contextual information is given to autonomous agents using the UI.