The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill assembles a shell command string dynamically in Step 2: caffeinate -is /bin/sh -c 'sleep <SLEEP_SECONDS> && cd "$HOME" && claude -p "Reply with exactly: ok"'. This command is executed in the background using calculated variables derived from user input.\n- [PRIVILEGE_ESCALATION]: In the Troubleshooting section, the skill explicitly suggests using dangerouslyDisableSandbox: true if the caffeinate utility is blocked by the execution environment. This flag bypasses platform-level security restrictions intended to limit the agent's access to the host system.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (e.g., 'remaining time until reset') to calculate the SLEEP_SECONDS value. If this input is not properly sanitized, an attacker could attempt command injection (e.g., by providing a time string like 1; rm -rf /) which would then be interpolated into the shell command string.\n- [COMMAND_EXECUTION]: The skill uses multiple system-level shell commands including ps, grep, date, caffeinate, sleep, and pkill to monitor and manage local processes on the user's machine.

claude-code-ping-start-5h-quota