The Agent Skills Directory

[SAFE]: The skill describes the use of official Atmos CLI commands for local infrastructure emulation. The functionality is consistent with the stated purpose of managing development environments.
[COMMAND_EXECUTION]: The skill instructions include commands for container management and execution, such as atmos emulator up and atmos emulator exec. These are legitimate administrative commands for the Atmos tool, which is authored by the same vendor (Cloud Posse) as the skill itself.
[INDIRECT_PROMPT_INJECTION]: The skill processes stack configuration and interacts with external container runtimes, representing an attack surface for malformed inputs.
Ingestion points: Reads Atmos stack configuration files (YAML) to resolve emulator settings and connection details.
Boundary markers: None explicitly mentioned in the documentation for separating configuration from command execution logic.
Capability inventory: Includes the ability to start, stop, reset, and execute arbitrary commands within emulator containers via the atmos CLI.
Sanitization: Relies on the underlying Atmos CLI's internal validation for configuration parsing and command execution.

atmos-emulator