The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches configuration and tooling from official vendor repositories and registries (e.g., @railcode/sdk, @railcode/next, and the railcode CLI). These are trusted resources necessary for platform functionality.
[COMMAND_EXECUTION]: The instructions guide the agent to use standard development commands, such as npm install, npx, and the platform-specific railcode CLI, for managing project lifecycle, builds, and deployments.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze existing project source code (e.g., Dockerfile, package.json, .env files) to generate a migration plan. While this ingestion of external data is a potential attack surface, the skill includes explicit guidance to verify all mappings with the user and emphasizes secure handling of sensitive information.
[DATA_EXFILTRATION]: No evidence of unauthorized data exfiltration was found. The skill actively promotes the use of secure platform features for secret management and egress control to handle credentials and outbound network requests.
[DYNAMIC_EXECUTION]: The provided build scripts (railcode/build.mjs) use standard Node.js subprocess execution to coordinate framework builds and deployments. These scripts are transparently documented as part of the migration workflow.

migrate-to-railcode