The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/deploy.sh script utilizes eval to execute the MIGRATIONS_CMD environment variable. While this is a common and intended feature for running database migrations during deployment, it involves executing content provided via the environment.
[DYNAMIC_EXECUTION]: The scripts/preflight.sh script sources a local .env.prod file using the source command to load environment variables. This is a standard functional pattern for deployment tools to validate required configuration before proceeding.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local data through several ingestion points.
Ingestion points: scripts/preflight.sh reads and parses .env.prod, Dockerfile, and docker-compose.prod.yml.
Boundary markers: The scripts do not use explicit delimiters or "ignore instructions" markers when reading these files, relying on standard shell parsing.
Capability inventory: The skill possesses significant capabilities, including full shell execution (bash), container management (docker, docker compose), and network requests (curl).
Sanitization: Validation is performed via docker compose config --quiet and tool availability checks, though user-defined environment variables used in eval are not sanitized.

dockerize-and-deploy