The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for tools that ingest external, potentially untrusted data which could contain malicious instructions.
Ingestion points: The n8n_deploy_template tool fetches workflow definitions from a remote template library, and get_node with includeExamples: true retrieves external configuration examples.
Boundary markers: The skill contains explicit 'side-effect boundaries' and warnings advising that templates are untrusted third-party code and should be reviewed before deployment.
Capability inventory: The toolset includes n8n_create_workflow, n8n_update_partial_workflow, and n8n_deploy_template, which provide significant write and execution capabilities on a live n8n instance.
Sanitization: The documentation recommends reviewing diffs before mutation, using synthetic staging credentials, and disabling automatic upgrades until compatibility is verified.
[DYNAMIC_EXECUTION]: The skill supports the creation and management of n8n workflows that include Code nodes, which execute arbitrary JavaScript or Python logic at runtime.
[EXTERNAL_DOWNLOADS]: The n8n_deploy_template tool is designed to download and deploy workflow configurations from an external library containing over 2,700 templates.

n8n-mcp-tools-expert