The system constructs working directories for each job, so that each run has a labeled and isolated area in which to work. It makes the working directory names using the values provided for the server path, project name, and tag.
When it executes a command, the system starts the command in the directory specified for the step. By default, that directory is the job's working directory, but you can also specify some other directory relative to the server's Path property. The topics in this section describe the path and directory creation processes in more detail.