The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Railway project configurations (such as project names and descriptions) which are controlled by users and potentially third parties.
Ingestion points: Results from railway list --json and railway status --json are used to provide project summaries.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the ingested project metadata.
Capability inventory: The skill can execute Railway CLI commands and call an internal API library to modify infrastructure settings.
Sanitization: The skill does not explicitly sanitize or escape project metadata before the agent processes and summarizes it.
[COMMAND_EXECUTION]: The skill executes shell commands using the Railway CLI and a local shell script (railway-api.sh) to interact with the Railway API.

railway-projects