Guides

Each guide covers one job you actually do with tomo-labs, grounded in the real commands and files. They assume you have worked through the quick start and have at least one tool built and run. They are ordered the way you tend to reach for them: understand the machine, read what it produced, look at what it tested, then extend and maintain it.

• How it works walks a run end to end: the proxy, the worker pool, and what keeps a rerun meaning the same thing. Read this first if you want to trust the numbers before you read them.
• Results is the current comparison table across all eight wired tools, and the 00-hello baseline that isolates each tool's fixed round-trip cost.
• Scenarios lists every task in the suite, from a baseline greeting through a small project scaffold.
• Adding a tool covers the two files a new agent needs to join the comparison.
• Upgrading the tools covers keeping the wired agents current as they ship new versions, and refreshing the captured results and prompts after a bump.

Beyond the core scenarios, the evals section covers the eval tiers: whole public benchmarks rendered into the same task shape and selected with --suite.