The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents how to execute the nxc and cme command-line tools to perform network-wide operations, including scanning subnets, verifying credentials, and executing commands on remote targets via protocols such as SMB, WinRM, and MSSQL.
[PRIVILEGE_ESCALATION]: The skill includes command examples that modify system configuration files, such as /etc/hosts and /etc/krb5.conf. While these are features of the tools used for domain resolution setup, they involve modifying sensitive system paths.
[INDIRECT_PROMPT_INJECTION]: The skill exposes an attack surface for indirect prompt injection by instructing the agent to process data from untrusted network sources.
Ingestion points: Network service banners, hostnames, SMB share names, and LDAP directory objects returned by the auditing tools (detected in SKILL.md).
Boundary markers: The instructions do not define boundary markers or directives for the agent to ignore instructions embedded in the tool output.
Capability inventory: The skill exhibits capabilities for network scanning, remote command execution on targets, and local file writing for configuration or reporting (SKILL.md).
Sanitization: There are no documented sanitization or validation steps for the data retrieved from the network.

nxc-cme