The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses tmux send-keys to execute commands in terminal sessions. While this is the intended functionality for an interactive TTY skill, it grants the agent the ability to run arbitrary shell commands.
[INDIRECT_PROMPT_INJECTION]: The skill captures terminal output using tmux capture-pane (utilized in SKILL.md and scripts/wait-for-text.sh) and passes it back to the agent. If the process running in the terminal outputs malicious instructions, the agent could potentially be influenced to perform unintended actions.
Ingestion points: Terminal output is ingested via tmux capture-pane in SKILL.md and the polling loop in scripts/wait-for-text.sh.
Boundary markers: The provided scripts do not implement delimiters or safety instructions to wrap the terminal data before processing.
Capability inventory: The skill can execute commands via tmux send-keys and manipulate the terminal server via kill-session and kill-server commands.
Sanitization: There is no sanitization or escaping of terminal output before it is presented to the agent's context.

tmux