The Agent Skills Directory

[SAFE]: No malicious behaviors or patterns were detected. The skill's capabilities are entirely consistent with its stated purpose of managing network devices within a Cisco Modeling Labs (CML) environment.
[COMMAND_EXECUTION]: The tool execute_command allows users to run CLI commands on managed virtual network nodes. This is an intended function for network automation and is restricted to the simulated lab environment.
[DATA_EXFILTRATION]: Configuration tools such as get_node_config and download_lab_configs are used to retrieve network device configurations for management and review, which is standard for this use case.
[PROMPT_INJECTION]: The skill processes external data from node console logs and command outputs, creating an indirect prompt injection surface.
Ingestion points: get_node_console_log, execute_command, and get_node_config as defined in SKILL.md.
Boundary markers: Absent in the instructional templates.
Capability inventory: Node state control, configuration management, and CLI execution tools listed in SKILL.md.
Sanitization: Absent.
Context: This surface is low-risk and inherent to the functionality of processing device outputs in a technical environment.

cml-node-operations