Validation framework
All agent actions inherit a validation system. Each agent action can include some validation steps using JSON schema or create custom validation steps.
JSON schema
Basic validation of input and output can be enforced using JSON Schema.
To validate the input just drop a validate-input.json file inside the
action directory, to validate the output just create a
validate-output.json file.
As general rule, schema files are loaded from special paths:
${AGENT_INSTALL_DIR}/validator-definitions.json: this is always picked up by the agent from all actions. It can contain the common data format definitions${AGENT_INSTALL_DIR}/actions/<action-name>/validate-input.json: schema applied to validate the input data${AGENT_INSTALL_DIR}/actions/<action-name>/validate-output.json: schema applied to validate the output data
Custom validation
The JSON schema can perform formal validation of input and output parameters. Semantic validation, like searching for duplicates or check if a network host is reachable, should be implemented inside custom scripts.
Additional validation logic can be implemented in an early step script
like 01validation. If the validation fails, the step must execute the
set-status command and return a non-zero exit code.
It is recommended to write custom scripts using Python, but any executable script inside the action directory would do the job.
Progress handling
Validations steps will block the task progress to 0 inside the UI. After the validation has passed, the task will be set as running and the progress will be tracked by the user interface.
When writing Python-based validation scripts, use this snippet:
agent.set_weight(os.path.basename(__file__), 0) # Validation step, no task progress at all