The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes internal shell scripts (spawn.sh, status.sh, check.sh) and the tmux system utility to manage background coding processes.
[EXTERNAL_DOWNLOADS]: The skill configuration includes a step to install the tmux binary using the Homebrew (brew) package manager during the installation process.
[PROMPT_INJECTION]: The skill acts as a wrapper for local and cloud-based AI agents and accepts user-defined tasks that are passed to these agents, creating a surface for indirect prompt injection. * Ingestion points: User input provided via the <task> argument in the spawn.sh command. * Boundary markers: The documentation does not define any delimiters or instructions to isolate user input from the agent's core instructions. * Capability inventory: The agents managed by this skill can perform code modifications, refactoring, and research, and can interact with local systems via Ollama or cloud APIs. * Sanitization: No sanitization or validation of the input task is described in the provided skill file.

tmux-agents