The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides several command-line examples for executing nxc (NetExec) and manspider to interact with network services and extract system secrets.
Evidence: Includes commands for SMB authentication checks, RID cycling, share spidering, and dumping SAM/LSA/NTDS secrets from target hosts.
[INDIRECT_PROMPT_INJECTION]: The skill documents modules designed to ingest and process data from external, untrusted sources (SMB shares), which creates a vulnerability surface where malicious content on those shares could influence the agent's behavior.
Ingestion points: Remote network shares crawled by the spider_plus module and document contents parsed by the manspider tool.
Boundary markers: No explicit delimiters or instructions to ignore embedded content are provided in the documented command workflows.
Capability inventory: The skill commands have the capability to perform network scanning, credential validation, and secret extraction (--sam, --lsa, --ntds).
Sanitization: No evidence of content sanitization or validation is present in the skill's instructions for handling results from the spidering process.

crackmapexec