The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documents installation procedures that download and execute scripts directly from the vendor's official domains (install.loops.so and github.com/loops-so).
[COMMAND_EXECUTION]: Provides instructions for the agent to execute the loops CLI locally to perform operational tasks, manage API keys, and interact with the Loops service.
[INDIRECT_PROMPT_INJECTION]: The skill uses local files (JSON property files and LMX markup files) as arguments for CLI commands. This constitutes an attack surface where maliciously crafted local data could influence command execution.
Ingestion points: File paths passed to flags such as --contact-props, --props, --json-vars, and --lmx-file in references/cli.md.
Boundary markers: No explicit markers or warnings are provided for the contents of these external files.
Capability inventory: The skill uses subprocess execution of the loops binary to send data to external APIs.
Sanitization: There is no mention of validating or sanitizing the content of the referenced local files before they are passed to the CLI.
[EXTERNAL_DOWNLOADS]: References downloading the CLI software from official sources including Homebrew, Go package registries, and the vendor's GitHub repository.

loops-cli