The Agent Skills Directory

[COMMAND_EXECUTION]: The railway_run tool enables the execution of arbitrary Railway CLI subcommands. This provides a significant attack surface, as the CLI is highly capable and could be used to perform actions outside the intended scope of the skill if the input is not strictly controlled.
[DATA_EXFILTRATION]: The skill provides tools like railway_logs, railway_build_logs, and railway_run (specifically for 'variables list') which can expose sensitive information. This includes environment variables (often containing API keys and database credentials) and application logs which may contain sensitive data or internal system details.
[EXTERNAL_DOWNLOADS]: The skill recommends installing the official @railway/cli via pnpm or brew. While these are legitimate tools for the Railway platform, they represent an external dependency required for the skill to function.
[PROMPT_INJECTION]: The skill processes untrusted data from deployment and build logs, making it susceptible to indirect prompt injection.
Ingestion points: Deployment and build logs fetched from the Railway platform via railway_logs and railway_build_logs.
Boundary markers: No specific boundary markers or isolation instructions are provided to help the agent distinguish between log data and instructions.
Capability inventory: The skill can execute CLI commands, modify deployments, and read sensitive environment variables.
Sanitization: Log content is processed and presented to the agent without any sanitization or filtering.

railway