What can you do with it?

Access and process output files (artifacts) from previous automation steps. Read data, extract content from documents, combine results, and work with various file formats seamlessly. Essential for multi-step workflows where data needs to be passed between steps.

How to use it?

Basic Command Structure

/artifacts [action] [file]

Parameters

Required:

  • action - What to do (get, list, extract, combine)

  • file - Artifact filename or pattern

Optional:

  • step - Step ID (like ‘a3x7’) or step number

  • property - Specific JSON property to extract

  • format - How to process the file

Response Format

The command returns:

{
  "data": "requested artifact content",
  "metadata": {
    "filename": "artifact name",
    "step": "source step ID",
    "type": "file type"
  }
}

Examples

Basic Usage

/artifacts
action: get data from step a3x7
file: results.json

Get data from a specific previous step by ID.

Advanced Usage

/artifacts
action: get user email from previous step
file: user-data.json
step: b2y9
property: contact.email

Extract specific property from JSON artifact in a previous step.

Specific Use Case

/artifacts
action: combine all CSV files into one dataset
file pattern: *.csv
steps: all previous steps

Combine multiple CSV artifacts from all previous steps into one dataset.

Notes

Automatically parses JSON and CSV files. Extracts text from PDFs, Word docs, Excel, and PowerPoint. Supports regex patterns for finding multiple files. Returns null if artifact not found. Step IDs (like ‘a3x7’) are more stable than step numbers.