The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple bash scripts (list-machines.sh, audit-machine.sh) to perform its primary function. These scripts are located in a hardcoded path on the host system within the author's home directory (/home/daniel/repos/github/my-repos/Claude-Security-Auditor-Plugin).
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface (Category 8) because it ingests and processes untrusted data from machine lists to drive shell execution. 1. Ingestion points: Data enters the context via the JSON output of the list-machines.sh script. 2. Boundary markers: Absent; there are no instructions to the agent to ignore potentially malicious content within the machine names. 3. Capability inventory: The skill has the capability to execute bash scripts (audit-machine.sh) with arguments and navigate the file system. 4. Sanitization: Absent; while the variable is quoted in the loop, there is no explicit validation or filtering of the machine names before they are used as command-line arguments.

audit-all