The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external Kedro project files such as catalog.yml, parameters.yml, and Python node files. This creates a surface where malicious instructions embedded in those files could influence the agent's behavior during the migration process.
Ingestion points: Kedro project files including catalog.yml, parameters.yml, and node code (Phase 1).
Boundary markers: The skill does not define specific delimiters or warnings to ignore instructions within the processed data.
Capability inventory: The agent is authorized to generate new Python scripts and project structures, and is instructed to run the zenml init shell command.
Sanitization: There is no explicit instruction to sanitize or filter the content of the Kedro files before they are processed by the LLM.
[DATA_EXFILTRATION]: The skill instructs the agent to read sensitive configuration files, specifically conf/local/credentials.yml, as part of the analysis phase. While this is necessary for migrating authentication logic, it involves the handling of potentially sensitive data.
Note: The skill includes explicit best-practice guidance to move these credentials into ZenML secrets or service connectors rather than inlining them into code, which mitigates the risk of exposure in the target project.
[COMMAND_EXECUTION]: The workflow requires the agent to execute shell commands, specifically zenml init, to set up the new project environment.
[EXTERNAL_DOWNLOADS]: The skill recommends that the user add a Model Context Protocol (MCP) server from https://docs.zenml.io/~gitbook/mcp. This is a trusted external resource belonging to the official vendor domain.

kedro-to-zenml-migration