Ask Claude Code a question headlessly and capture the answer in
/root/claude-answer.txt.
claude -p "Explain the difference between a process and a thread in two sentences." > /root/claude-answer.txt
cat /root/claude-answer.txt
That's the whole pattern: -p makes Claude answer once and exit, stdout is
the answer, and the shell does the rest.
A two-sentence explanation sitting in a file — produced by an AI agent you invoked like any other Unix tool.
claude -p "..." | wc -w tell you?Press Submit when the file has content.
Start the lab on the right to run checks.