Custom Functions automation integration with GenticFlow
Admin-defined Python functions GenticFlow can call under policy
GenticFlow custom functions let admins define named Python functions with descriptions and code snippets, validate them against blocked patterns, tag them, and attach them to GenticFlow workflows. Functions extend what GenticFlow can call without hard-coding every customer-specific operation into the product.
What GenticFlow does with this integration
Function Management
- Functions have a name, description, function_name, and Python code snippet
- Custom and built-in system functions are listed together for selection
- Functions can be tagged for organization
- GenticFlow workflows can be configured with selected functions as a data source
Validation and Execution Controls
- Function code is validated before save unless explicitly force-saved by an authorized admin
- The executor applies module and operation restrictions before running a function
- Function definitions must expose the expected top-level function name
- Execution follows the backend's controlled FunctionExecutor paths
Extensible Automation
- Use custom functions for customer-specific lookups, transformations, or controlled API calls
- Function Assistant can help generate function code in the UI
- Credential references can be used in generated function guidance
- Admins keep ownership of what callable tools are exposed to agents
How It Works
Create a custom function
Define the display name, description, function_name, and Python code snippet.
Validate the implementation
GenticFlow checks the function definition and flags blocked code patterns before save.
Attach it to a workflow
Select custom or built-in functions for the GenticFlow workflow that should be allowed to call them.
Use it during automation
The workflow can call the exposed function when its task requires that customer-specific capability.
Automation with guardrails.
Custom Functions extends GenticFlow without bypassing the product controls. Events, alerts, or approved callable functions can enter the workflow while validation, permissions, and execution history stay in place.
Request a Demo