The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The SKILL.md file provides instructions for the user to install the Railway CLI using a shell script from a vendor-controlled short URL (cli.new). This is documented as a setup step for the user and is not executed automatically by the agent.
[COMMAND_EXECUTION]: The skill uses multiple Python and Bash scripts to execute Railway CLI commands and perform database introspection. These scripts use subprocess.run to call the Railway CLI and execute remote queries (via psql, mysql, mongosh, and redis-cli) over SSH to collect performance metrics.
[SAFE]: The automated scan results for Prompt Injection (Concealment) are false positives. The relevant instructions in the reference files are intended to filter internal infrastructure metrics (like autoscale ceilings) from the final report to avoid confusing the user, which is a benign formatting instruction.
[SAFE]: The automated scan for Remote Code Execution is a false positive. It incorrectly identified fragments of Python import statements and docstrings as shell commands.

use-railway