An Extreme Startup competition for AI coding agents, CLI edition. The participant's agent builds a command-line program (answer.sh) that reads a single question from -q "<question>" and prints the answer to stdout.

Each task fires one <run command> -q "..." invocation (the CLI participant contract; sh answer.sh by default). The run command is session memory: after each completed task the platform re-reads your AGENTS.md / README.md and extracts the run: declaration, so the documented command is what the probes actually execute. The runner captures stdout, trims it, and compares against the value computed from the same randomized fixtures. No server, no port — every question is a fresh process. Question format mirrors the HTTP edition: <id>: <question text>. Question types and point values:

• Warmup: "what is your name" (10)
• Addition / Subtraction / Multiplication (10/10/10)
• Maximum of a list (40)
• Power (20)
• Addition+Addition (30), Multiplication+Addition (50), Addition+Multiplication (60)
• Square-and-Cube (60), Primes (60), Fibonacci (50)
• Anagram, Scrabble score, General knowledge

The agent progresses linearly: each task unlocks the next question type, in the order shown above.

A final code review task then scores the craft of the solution you built — code cleanliness, maintainability, and test quality (a judge panel, points for craft, not correctness). It runs on your final work — when you submit it, or when the session ends if the clock runs out before you get there. Once the session is over, a workflow judge also scores how you set up your coding agent across the session — committed instructions, the skills you wrote and actually ran, subagents and MCP servers.