Workflows (Beta)

Workflow Blocks

TaskBlock

The magic block. Skyvern navigates through the websites to take actions and/or extract information.

Example block:

- block_type: task
      label: login
      parameter_keys:
        - credentials
      url: website_url
      navigation_goal: >-
        If you're not on the login page, navigate to login page and login using
        the credentials given. First, take actions on promotional popups or cookie prompts that could prevent taking other action on the web page. If you fail to login to find the login page or can't login after several trials, terminate. If login is
        completed, you're successful. 
      data_extraction_goal: >-
        Extract anything for the sake of this demo
      error_code_mapping:
        stuck_with_popups: terminate and return this error if you can't close popups after several tries and can't take the necessary actions on the website because there is a blocking popup on the page
        failed_to_login: terminate and return this error if you fail logging in to the page

Inputs:

  1. URL (often required): Skyvern Agent’s starting point. Ideally the website you’d like to automate.
    • In the workflows interface, if this input is left blank it will continue where the previous node left off. The idea of a navigation goal is to set or reset where the agent starts off.
    • If you’re logging in to a site using the first task block, you would want to leave URL blank for the second block so that it can continue after logging in
  2. Navigation Goal (often required): details where Skyvern is going and what Skyvern is doing. Clear Navigation Goals will be a single goal, broken down into steps. Avoid supplying multiple goals. You need to specify when the goal is complete, using “COMPLETE”, or when to abandon that goal, using “TERMINATE”
    • The navigation goal is not used to load the URL. Asking Skyvern to “go to website A” in this field will not have the intended effect
    • Terminations result in Skyvern explaining why it stopped navigating
    • This field can be omitted if you only want Skyvern to extract data without navigating anywhere else
  3. Data Extraction Goal (optional): aside from where Skyvern is going and what Skyvern is doing, is there anything that Skyvern is extracting and returning back? A good data extraction goal is specific about what Skyvern is returning to the user
    • Note that data extractions only happen after Skyvern is finished navigating
  4. Extracted Information Schema (optional): if you have a data extraction goal, some users need it formatted in a certain way for internal purposes. Navigation payload accepts JSON formatted specifications for how the data should be returned
  5. Max Steps Override (optional): some users want to cap cost through the number of steps the task can take
  6. Max Retries (optional): the number of times a step can be retried if it fails
  7. Complete on Download (optional): Allows Skyvern to complete the task after a file has been downloaded
  8. File Suffix (optional): an identifier attached to the downloaded file
  9. TOTP URL and TOTP Identifier (optional): if you have an internal system where you can store the 2FA TOTP code, this URL calls on this storage space. The identifier allows you to link the code to the task, this is critical if you are running multiple tasks concurrently. Contact us if you would like to set up 2FA retreival in your workflows.
  10. Parameters (optional): parameters are self-defined placeholders that are specified run-to-run. They can either be workflow parameters, passed in via an API call, or output parameters, extracted from a previous task block. If specified, they are used by Skyvern to assist in the navigation, to fill out forms or further influence what actions to take on a website.

ForLoopBlock

Iterate over something such as a CSV or the output of a previous block. The blocks nested under loop_blocks are the blocks that will be repeated for each entry in the

- block_type: for_loop
      label: iterate_over_order_ids
      loop_over_parameter_key: order_ids
      continue_on_failure: true
      loop_blocks:
        - block_type: task
          label: download_invoice_for_order
          complete_on_download: true
          continue_on_failure: true
          parameter_keys:
            - order_id
          url: order_history_url
          navigation_goal: >-
            Download the invoice of the order with the given order ID. 
            Make sure to download the invoice for the given order id. 
            If the element tree doesn't have a matching order id, check the screenshots. 
            Complete if you have successfully downloaded the invoice according to action history, if you were able to download it, you'll see download_triggered=True for the last step. 
            If you don't see a way to download an invoice, navigate to the order page if possible. 
            If there's no way to download an invoice terminate. 
            If the text suggests printing, you can assume you can download it. 
            Return click action with download=True if you want to trigger a download.
          error_code_mapping:
            not_possible_to_download_invoice: return this error if the website doesn't allow downloading/viewing invoices
            cant_solve_captcha: return this error if captcha isn't solved after multiple retries

Inputs:

  1. Loop Value (required): This is the value that the loop will iterate over. For instance, if you have for every invoice ID, do X, invoice ID would be the value for this input.
    • Please contact us if you would like to add a loop block. Since we’re in beta, the loop value needs to be parameterized from the backend.
  2. Another block nested within the loop (required)

CodeBlock

This block executes user-defined Python code within our execution environment. It’s able to take parameters as input and transform them based on a certain specification.

In addition to running simple code snippets, CodeBlock allows you to:

  • execute asynchronous code
  • control your browser page inside Skyvern

Example Block

1- block_type: code
2  label: calculate_percentage_diff
3  parameter_keys:
4    - alibaba_price
5    - amazon_price
6  code: |
7    if amazon_price["unitPrice"] and alibaba_price["unitPrice"]:
8        result = 1.0 * (alibaba_price["unitPrice"] - amazon_price["unitPrice"]) / amazon_price["unitPrice"]
9    else:
10        result = None
11  output_parameter_key: price_diff_percentage

Example Block with Browser Control

1- block_type: code
2  label: get_tab_details
3  code: |
4    print("Getting tab details")
5    result = {
6        "url": skyvern_page.url,
7        "title": await skyvern_page.title()
8    }
9    print("Got details:", result)
10    print("Now I want to see a cat")
11    await skyvern_page.goto("https://cataas.com/cat")

Inputs:

  1. Code (required): Insert your custom Python code so that you can define your own custom block.

TextPromptBlock

Do a custom OpenAI query as a part of your workflow

- block_type: text_prompt
  label: generate_new_title
  parameter_keys:
    - alibaba_title
    - amazon_title
  llm_key: OPENAI_GPT4O
  prompt: >
    You're given two e-commerce product titles. Use both and generate a
    better one.
      Title 1: {{ alibaba_title }}
      Title 2: {{ amazon_title }}
  output_parameter_key: new_title

Inputs:

  1. Prompt (required): Write a natural language prompt to be sent to the LLM to generate a text response
  2. JSON Schema (optional): Craft a JSON input that structures the LLM output for use in another programming task

FileUploadBlock

Persists files inside custom destinations. Supported integrations:

  • AWS S3
- block_type: file_upload
  label: upload_downloaded_files_to_s3
  storage_type: s3
  aws_access_key_id: YOUR_AWS_ACCESS_KEY_ID
  aws_secret_access_key: YOUR_AWS_SECRET_ACCESS_KEY
  s3_bucket: YOUR_S3_BUCKET
  region_name: us-east-1

How to set up FileUploadBlock with AWS

  • Step 1. Create your bucket. (let’s say YOUR-BUCKET-NAME is the name)
  • Step 2. Create a limited IAM policy that only has GetObject and PutObject permissions to this s3 bucket, named skyvern-s3-access-policy (any name works): 
    1{
    2    "Version": "2012-10-17",
    3    "Statement": [
    4        {
    5            "Sid": "SkyvernS3Access",
    6            "Effect": "Allow",
    7            "Action": [
    8                "s3:PutObject",
    9                "s3:GetObject"
    10            ],
    11            "Resource": "arn:aws:s3:::YOUR-BUCKET-NAME/*"
    12        }
    13    ]
    14}
  • Step 3. Create an AWS IAM user for skyvern
    • Go to https://us-east-1.console.aws.amazon.com/iam/home
    • click “Create User” → do not check “Provide user access to the AWS Management Console” → input user name → click “Next”
    • At the the “Set Permissions” page, click “Attach policies directly” → pick the skyvern-s3-access-policy created in the previous step → click “Next” → click “Create user”
    • After the user is created, go to the new IAM user page → go to the “Security credentials” tab → click “create access key” → on the “Access key best practices & alternatives” page, click “Application running on an AWS compute service” → click “Next” → click “Create access key” → you will see the “Access Key” and the “Secret access key” at the end of user creation.
  • Step 4. Create a FileUploadBlock
    • Place this block after the file download has completed.
    • Copy the “Access Key” to the “AWS Access Key ID” field of the File Upload Block and “Secret access key”.
    • Copy the “Access Key” to the “AWS Secret Access Key” field in the File Upload Block.
    • Add your s3 bucket name.
    • Add your AWS region.

SendEmailBlock

Sends an email with some data

- block_type: send_email
  label: send_email
  smtp_host_secret_parameter_key: smtp_host
  smtp_port_secret_parameter_key: smtp_port
  smtp_username_secret_parameter_key: smtp_username
  smtp_password_secret_parameter_key: smtp_password
  sender: hello@skyvern.com
  recipients:
    - founders@skyvern.com
  subject: Skyvern - Downloaded Invoices Demo
  body: website_url
  file_attachments:
    - SKYVERN_DOWNLOAD_DIRECTORY

Inputs:

  1. Recipients (required): a list of people who will receive the email separated by commas
  2. Subject/Body (optional): the header and body of an email
  3. File attachments (optional): since we’re still in beta, you will need to contact us to upload attachments to the email

FileParserBlock

Downloads and parses a file to be used within other workflow blocks.

Supported types: CSV

- block_type: file_url_parser
  label: csv_parser
  file_type: csv
  file_url: <csv_file_url>

Inputs:

  1. File URL (required): This block allows you to use a CSV within your workflow.
    • Since we’re still in beta, you will need to contact us to load a value into this block