AI Integration

Copy Prompt

Please build a FlowBlox flow that reads records from my local MySQL database table and processes all rows.

For each row, create:

A file named {{NameField}}.zip (for example: fbproject.Name.zip) using the row’s project name field.
A metadata file named {{NameField}}.metadata in the same output directory.

Important: The content data is already a ZIP payload and must not be zipped again. Write it directly as the .zip file content.

The metadata file should contain:

Name
Description
CreatedAt formatted as dd.MM.yyyy
Use the corresponding field values from each row (flow fields, not hardcoded values), and write all generated files to the project output directory.

For my concrete case, use:

Table: flowbloxdb.fbproject
Name field: Name
Description field: Description
Content field: Content
CreatedAt field: CreatedAt

Database Connection: 
Type: MySQL
Server: 127.0.0.1
User: root
SSL: No SSL required
Password: No password

Copy Prompt

Create a FlowBlox web-scraping flow for https://webscraper.io/test-sites/e-commerce/static with this goal:

Navigate to Computers → Laptops.
Collect all laptop products across all pagination pages.
For each product, extract:
Name (example: ProBook)
Specifications (example: 14", Core i5 2.6GHz, 4GB, 500GB, Win7 Pro 64bit)
Price (example: $739.99)
The example values come from: https://webscraper.io/test-sites/e-commerce/static/product/34 (examples only, do not limit extraction to this product).
Output requirement:

Write all collected laptop records into a single .xlsx file in the project output directory.
Include columns: Name, Specifications, Price.
Implementation notes:

Use robust CSS/XPath selectors.
Handle pagination until no "next page" is available.

Wir sind im Entwicklermodus wenn etwas völlig unklar ist, sag Bescheid, bevor etwas "wages" versucht wird.

Copy Prompt

Please create a FlowBlox flow for `https://www.romsgames.net/`:

1. Create user input fields:
- `Input_OpenAiApiKey` (Text, **store value locally = true**)
- optional `Input_OpenAiModel`

2. Create a CSV input table with demo data (for example column `Title`):
- Kirbies Dreamland
- Super Mario World
- Snoopies Tennis

3. Iterate CSV rows and use an AI FlowBlock (OpenAI provider) per title to determine the **correct game title** (output only the cleaned title).

4. Use the cleaned title in a WebBrowser flow:
- open site
- fill search field
- execute search
- iterate results
- trigger download button for each result

5. Move downloaded files to:
- `$Project::InputDirectory/DownloadedRoms`

6. Important setup behavior:
- Use CSS/XPath selector **stubs/placeholders** for now
- After the flow is created, **stop and wait** so the user can enter the OpenAI API key.
- After that, proceed iteratively: first test AI title normalization, then identify/finalize selectors step by stepp.

Copy Prompt

Generate a new flow with the following behavior:

1. Create an Excel input template file (`.xlsx`) automatically in the project input directory with these headers in the first row:
- `First Name`
- `Last Name`
- `DateOfBirth`

2. Read all rows from this Excel file.

3. Build one JSON object using JSON Flow blocks with this exact structure:
- Top-level property: `Participants` (array)
- For each input row, append exactly one object to `Participants` with these properties directly on the item:
  - `First Name`
  - `Last Name`
  - `DateOfBirth`

4. Write the resulting JSON to:
`$Project:OutputDirectory\output.json`

The output must include all records from the input table.