vodorod/Dorod-Sky
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Workflow Copilot: some improvements (#4413)

Browse Source
This commit is contained in:
Stanislav Novosad
2026-01-07 20:47:27 -07:00
committed by GitHub
parent 66d28bb24d
commit a1f0adf273
2 changed files with 253 additions and 35 deletions
Download Patch File Download Diff File Expand all files Collapse all files

286
skyvern/forge/prompts/skyvern/workflow_knowledge_base.txt
View File

@@ -93,18 +93,16 @@ Important Rules:
- Set next_block_label to null to mark the end of a flow
- continue_on_failure allows graceful error handling
** TASK BLOCK (task) **
** NAVIGATION BLOCK (navigation) **
Purpose: Navigate to a URL, perform actions based on natural language goals, and optionally extract data.
Purpose: Take actions to achieve a task. This is the "Browser Task" block in the UI.
Structure:
block_type: task
block_type: navigation
label: <unique_label>
url: <starting_url> # Optional: URL to navigate to; omit to continue on current page
title: str # Required: The title of the block
navigation_goal: <action_description> # Optional: What actions to perform
data_extraction_goal: <extraction_description> # Optional: What data to extract
data_schema: <json_schema> # Optional: Schema for extracted data
navigation_goal: <action_description> # Required: What actions to perform
error_code_mapping: {} # Optional: Map errors to custom codes
max_retries: 0 # Optional: Number of retry attempts
max_steps_per_run: null # Optional: Limit steps per execution
@@ -121,23 +119,16 @@ include_action_history_in_verification: false # Optional: Include history in ve
Use Cases:
- Fill out forms on websites
- Navigate complex multi-step processes
- Extract structured data from pages
- Combine navigation and extraction in one step
- Prepare the page for a subsequent extraction block
- Execute focused browser tasks with clear completion criteria
Example:
blocks:
- block_type: task
label: search_and_extract
- block_type: navigation
label: search_and_open
next_block_label: null
url: "https://example.com/search"
navigation_goal: "Search for {{ query }} and click the first result"
data_extraction_goal: "Extract the product name, price, and availability"
data_schema:
type: object
properties:
name: {type: string}
price: {type: number}
available: {type: boolean}
parameter_keys:
- query
max_retries: 2
@@ -155,7 +146,7 @@ max_retries: 0 # Optional: Retry attempts
parameter_keys: [] # Optional: Parameters used
Use Cases:
- Jump to a known page before a task block
- Jump to a known page before other blocks
- Reset the browser state to a specific URL
- Split URL navigation from subsequent actions
@@ -200,7 +191,7 @@ blocks:
** TASK V2 BLOCK (task_v2) **
Purpose: Task block that can handle complex, multi-step workflows using a single natural language prompt. Can handle more complex scenarios than task blocks but may be slightly slower.
Purpose: Achieve complex tasks with deep thinking using a single natural language prompt. Can handle more complex scenarios than navigation blocks but may be slightly slower.
Structure:
block_type: task_v2
@@ -215,10 +206,10 @@ Use Cases:
- General-purpose automation with flexible requirements
- When you need to handle more complex scenarios and are okay with potentially slower execution
Differences from Task Block:
- Uses single "prompt" field instead of separate "navigation_goal" and "data_extraction_goal"
Differences from Navigation Block:
- Uses single "prompt" field instead of "navigation_goal"
- Can handle more complex scenarios and longer sequences of actions
- May be slightly slower than task blocks
- May be slightly slower than navigation blocks
- No data_schema (extraction format described in prompt)
- More flexible configuration
@@ -275,6 +266,32 @@ blocks:
complete_criterion: "Current URL is 'https://portal.example.com/dashboard'"
max_retries: 2
** VALIDATION BLOCK (validation) **
Purpose: Validate workflow state and decide whether to continue or terminate.
Structure:
block_type: validation
label: <unique_label>
complete_criterion: <success_condition> # Optional: Condition for success
terminate_criterion: <termination_condition> # Optional: Condition to stop workflow
error_code_mapping: {} # Optional: Map errors to custom codes
parameter_keys: [] # Optional: Parameters used in this block
disable_cache: false # Optional: Disable caching
Use Cases:
- Confirm a successful navigation or submission
- Stop the workflow when an error state appears
- Validate content on the page before extracting data
Example:
blocks:
- block_type: validation
label: verify_submission
next_block_label: null
complete_criterion: "Page contains 'Thank you for your submission'"
terminate_criterion: "Page contains 'Error' or 'Try again'"
** EXTRACTION BLOCK (extraction) **
Purpose: Extract structured data from the current page without navigation.
@@ -332,6 +349,204 @@ blocks:
rating: {type: number}
max_retries: 1
** FILE DOWNLOAD BLOCK (file_download) **
Purpose: Download files from a website. This is the "File Download Block" in the UI.
Structure:
block_type: file_download
label: <unique_label>
navigation_goal: <download_instruction> # Required: How to trigger the download
url: <starting_url> # Optional: URL to navigate to first
title: str # Optional: Title for the block
engine: skyvern_v1 # Optional: Run engine
error_code_mapping: {} # Optional: Map errors to custom codes
max_retries: 0 # Optional: Number of retry attempts
max_steps_per_run: null # Optional: Limit steps per execution
parameter_keys: [] # Optional: Parameters used in this block
download_suffix: null # Optional: Full filename for the download
totp_verification_url: null # Optional: TOTP verification URL
totp_identifier: null # Optional: TOTP identifier
disable_cache: false # Optional: Disable caching
download_timeout: null # Optional: Download timeout in seconds
Use Cases:
- Download invoices, receipts, or reports from a portal
- Export data as CSV or PDF from a dashboard
- Trigger file downloads that require navigation steps
Example:
blocks:
- block_type: file_download
label: download_report
next_block_label: null
url: "https://portal.example.com/reports"
navigation_goal: "Open the latest report and click Download as PDF"
download_suffix: "latest_report.pdf"
** CLOUD STORAGE BLOCK (file_upload) **
Purpose: Upload files to storage. This is the "Cloud Storage Block" in the UI.
Structure:
block_type: file_upload
label: <unique_label>
storage_type: <s3|azure> # Optional: Storage backend (default: s3)
s3_bucket: <bucket_name> # Optional: S3 bucket name
aws_access_key_id: <access_key_id> # Optional: AWS access key id
aws_secret_access_key: <secret_access_key> # Optional: AWS secret access key
region_name: <aws_region> # Optional: AWS region
azure_storage_account_name: <account_name> # Optional: Azure storage account
azure_storage_account_key: <account_key> # Optional: Azure storage account key
azure_blob_container_name: <container_name> # Optional: Azure blob container
azure_folder_path: <folder_path> # Optional: Azure folder path
path: <local_or_workspace_path> # Optional: File path to upload
Use Cases:
- Upload downloaded artifacts to S3
- Publish files to Azure Blob Storage
- Persist workflow outputs to storage
Example:
blocks:
- block_type: file_upload
label: upload_report
next_block_label: null
storage_type: s3
s3_bucket: "my-reports"
region_name: "us-west-2"
path: "/tmp/latest_report.pdf"
** FILE PARSER BLOCK (file_url_parser) **
Purpose: Parse PDFs, CSVs, and Excel files. This is the "File Parser Block" in the UI.
Structure:
block_type: file_url_parser
label: <unique_label>
file_url: <https_url_or_file_url> # Required: URL to the file
file_type: <pdf|csv|excel> # Required: File type
json_schema: <json_schema> # Optional: Structure of parsed output
Use Cases:
- Parse a PDF invoice into structured fields
- Extract rows from a CSV file
- Read Excel sheets for downstream processing
Example:
blocks:
- block_type: file_url_parser
label: parse_invoice
next_block_label: null
file_url: "https://example.com/invoice.pdf"
file_type: pdf
json_schema:
type: object
properties:
invoice_id: {type: string}
total: {type: number}
** SEND EMAIL BLOCK (send_email) **
Purpose: Send email notifications. This is the "Send Email Block" in the UI.
Structure:
block_type: send_email
label: <unique_label>
smtp_host_secret_parameter_key: <param_key> # Required: Secret parameter key for SMTP host
smtp_port_secret_parameter_key: <param_key> # Required: Secret parameter key for SMTP port
smtp_username_secret_parameter_key: <param_key> # Required: Secret parameter key for SMTP username
smtp_password_secret_parameter_key: <param_key> # Required: Secret parameter key for SMTP password
sender: <email_address> # Required: Sender email address
recipients: [<email_address>] # Required: Recipient list
subject: <subject_line> # Required: Email subject
body: <email_body> # Required: Email body
file_attachments: [<file_path>] # Optional: Local file paths to attach
Use Cases:
- Notify a team when a workflow completes
- Send extracted data summaries to stakeholders
- Email reports or attachments
Example:
blocks:
- block_type: send_email
label: notify_ops
next_block_label: null
smtp_host_secret_parameter_key: smtp_host
smtp_port_secret_parameter_key: smtp_port
smtp_username_secret_parameter_key: smtp_user
smtp_password_secret_parameter_key: smtp_pass
sender: "automation@example.com"
recipients: ["ops@example.com"]
subject: "Daily report ready"
body: "The latest report is ready for review."
file_attachments: ["/tmp/latest_report.pdf"]
** TEXT PROMPT BLOCK (text_prompt) **
Purpose: Process text with LLM. This is the "Text Prompt Block" in the UI.
Structure:
block_type: text_prompt
label: <unique_label>
llm_key: <llm_key> # Optional: Model key override
prompt: <text_prompt> # Required: Prompt to run
parameter_keys: [] # Optional: Parameters used in this block
json_schema: <json_schema> # Optional: Structured output schema
Use Cases:
- Summarize extracted text or documents
- Normalize free-form content into a schema
- Generate classifications or tags
Example:
blocks:
- block_type: text_prompt
label: summarize_notes
next_block_label: null
prompt: "Summarize these notes: {{ notes }}"
json_schema:
type: object
properties:
summary: {type: string}
action_items: {type: array, items: {type: string}}
parameter_keys: [notes]
** HTTP REQUEST BLOCK (http_request) **
Purpose: Make HTTP API calls. This is the "HTTP Request Block" in the UI.
Structure:
block_type: http_request
label: <unique_label>
method: <GET|POST|PUT|PATCH|DELETE> # Optional: HTTP method (default: GET)
url: <https_url> # Optional: Target URL
headers: {} # Optional: HTTP headers
body: {} # Optional: JSON body (dict)
files: {} # Optional: Multipart files mapping
timeout: 30 # Optional: Timeout in seconds
follow_redirects: true # Optional: Follow redirects
parameter_keys: [] # Optional: Parameters used in this block
Use Cases:
- Call third-party APIs for enrichment
- Post data to internal services
- Upload files via multipart requests
Example:
blocks:
- block_type: http_request
label: lookup_customer
next_block_label: null
method: "POST"
url: "https://api.example.com/customers/search"
headers:
Authorization: "Bearer {{ api_token }}"
body:
email: "{{ customer_email }}"
parameter_keys: [api_token, customer_email]
** PARAMETER TEMPLATING **
All string fields in blocks support Jinja2 templating to reference parameters.
@@ -395,9 +610,10 @@ blocks:
- block_type: extraction
label: step2
next_block_label: step3
- block_type: task
- block_type: task_v2
label: step3
next_block_label: null
prompt: "Complete the final step on the page"
# Executes: step1 → step2 → step3
Explicit Flow Control (Skip blocks):
@@ -406,9 +622,10 @@ blocks:
label: login
next_block_label: extract_data
url: "https://app.example.com/login"
- block_type: task
- block_type: task_v2
label: handle_error
next_block_label: null
prompt: "Handle the error state if it appears"
- block_type: extraction
label: extract_data
next_block_label: null
@@ -416,10 +633,11 @@ blocks:
Error Recovery Flow:
blocks:
- block_type: task
- block_type: task_v2
label: primary_task
next_block_label: verify_result
continue_on_failure: true
prompt: "Attempt the primary task on the page"
- block_type: validation
label: verify_result
next_block_label: null
@@ -454,7 +672,7 @@ blocks:
* Performance:
- Use disable_cache: true for dynamic content
- Set max_steps_per_run to prevent infinite loops
- Combine navigation and extraction in task blocks when possible
- Use task_v2 when a single prompt should include both actions and extraction requirements
* Security:
- Never hardcode credentials in workflows
@@ -477,7 +695,7 @@ blocks:
next_block_label: go_to_reports
url: "https://app.example.com/login"
parameter_keys: [my_credentials]
- block_type: task
- block_type: navigation
label: go_to_reports
next_block_label: get_report_data
navigation_goal: "Navigate to Reports section"
@@ -507,24 +725,24 @@ blocks:
label: open_form
next_block_label: fill_personal_info
url: "https://forms.example.com/application"
- block_type: task
- block_type: navigation
label: fill_personal_info
next_block_label: fill_address
navigation_goal: "Fill in name as {{ name }}, email as {{ email }}"
parameter_keys: [name, email]
- block_type: task
- block_type: navigation
label: fill_address
next_block_label: submit
navigation_goal: "Fill in address fields and click Continue"
parameter_keys: [address, city, zip]
- block_type: task
- block_type: navigation
label: submit
next_block_label: null
navigation_goal: "Review information and click Submit"
Pattern 4: Conditional Extraction
blocks:
- block_type: task
- block_type: navigation
label: search_product
next_block_label: check_availability
navigation_goal: "Search for {{ product }}"
@@ -536,7 +754,7 @@ blocks:
type: object
properties:
in_stock: {type: boolean}
- block_type: task
- block_type: navigation
label: add_to_cart
next_block_label: null
navigation_goal: "If product is in stock, add to cart"
@@ -552,7 +770,7 @@ Workflow-Level:
Block-Level:
- label is required and cannot be empty
- block_type must be a valid type
- For task blocks: either navigation_goal or data_extraction_goal should be present
- For navigation blocks: navigation_goal is required
- For extraction blocks: data_extraction_goal is required
- For action blocks: navigation_goal is required
- For login blocks: parameter_keys should include credentials
@@ -594,7 +812,7 @@ workflow_definition:
- account_creds
complete_criterion: "URL contains '/dashboard'"
- block_type: task
- block_type: navigation
label: search_and_filter
next_block_label: get_product_info
url: "https://shop.example.com/search"

2
skyvern/forge/sdk/routes/workflow_copilot.py
View File

@@ -44,7 +44,7 @@ async def _get_debug_artifact(organization_id: str, workflow_run_id: str) -> Art
artifacts = await app.DATABASE.get_artifacts_for_run(
run_id=workflow_run_id, organization_id=organization_id, artifact_types=[ArtifactType.VISIBLE_ELEMENTS_TREE]
)
return artifacts[0] if isinstance(artifacts, list) else None
return artifacts[0] if isinstance(artifacts, list) and artifacts else None
async def _get_debug_run_info(organization_id: str, workflow_run_id: str | None) -> RunInfo | None: