The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the itential-mcp Python package and its dependencies, including fastmcp, ipsdk, python-toon, and wsproto, from public package registries to enable communication with the Itential platform.
[COMMAND_EXECUTION]: Provides tools such as run_command, run_service, and apply_device_configuration which allow the agent to perform CLI operations and configuration deployments on managed network infrastructure as part of its automation workflow.
[CREDENTIALS_UNSAFE]: Relies on environment variables ITENTIAL_MCP_PLATFORM_USER and ITENTIAL_MCP_PLATFORM_PASSWORD to authenticate with the remote platform, which is a standard configuration method for such integrations.
[PROMPT_INJECTION]: Contains an indirect prompt injection surface through tools that ingest external structured data. Ingestion points: template parameter in render_template (Jinja2) and openapi_spec in create_integration_model. Boundary markers: None identified in the skill definition. Capability inventory: High-privilege actions including run_command and apply_device_configuration. Sanitization: No specific input validation or sanitization logic is described in the skill documentation.

itential-automation