The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves retrieving existing workflow definitions from the Scenario API to serve as structural templates. While this introduces a data ingestion surface for external logic, it is the intended use case for authoring and cloning workflows.
Ingestion points: Workflow records retrieved from the Scenario API via the workflow_get tool.
Boundary markers: No specific delimiters or boundary markers are defined in the instructions to separate external graph data from agent instructions.
Capability inventory: The agent utilizes workflow_create, workflow_update, and workflow_publish tools to modify and deploy executable workflows.
Sanitization: The instructions guide the agent to manually trim and inspect workflow data; no automated sanitization of ingested CEL expressions or prompt strings is implemented.
[COMMAND_EXECUTION]: The skill includes a utility Python script (scripts/fetch_workflow_examples.py) for maintainers to fetch featured examples. The script includes a defensive path sanitization step to prevent directory traversal during file writing.
[DATA_EXPOSURE]: The provided Python script manages authentication by reading SCENARIO_SDK_API_KEY and SCENARIO_SDK_API_SECRET from environment variables, adhering to standard secure practices for API credential management.

scenario-workflow-authoring