The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the railway CLI to perform deployment operations, manage services, and link project directories. This is standard functionality for the service.\n- [DATA_EXFILTRATION]: Documentation identifies ~/.railway/config.json as the storage location for authentication tokens required for GraphQL API access. This is a standard path for the Railway CLI and is used to facilitate authorized interactions with the platform.\n- [EXTERNAL_DOWNLOADS]: The skill recommends installing python-json-logger for structured logging. This is a well-known, legitimate package from the official Python package registry.\n- [INDIRECT_PROMPT_INJECTION]: The skill enables reading application logs via the railway logs command, which represents a potential surface for indirect injection of untrusted data into the agent context.\n
Ingestion points: Application log output retrieved via the railway logs command.\n
Boundary markers: Not present in the provided instructions.\n
Capability inventory: Full application lifecycle management via the railway CLI (init, up, ssh, variables) and GraphQL API access.\n
Sanitization: No explicit sanitization of log content is defined.

railway