run_command tool erroring

Problem

Since enabling running commands there's been a higher number of failed tool calls for running commands.

While run_command can be expected to fail at a higher degree than other commands due to it's flexibility, from a first glance some of these failures seem to be avoidable and should be investigated further, examples:

• Validation error: 1
• No valid CLI command being produced in a case where built in tools should likely have been used: 1
• Git tools not being available: 1

Desired outcome

Avoidable run_command errors no longer happen

Proposed Solution