For AI agents and LLMs: a machine-readable index is available at llms.txt. A plain-Markdown version of any documentation page is available by appending .md to its URL.
Skip to main content

/generate Command

Use /generate after exploration to write test scenarios for the active agent's discovered features.

Current Rook generate command help with total, class, category, force, allow, JSON, and verbose options

Syntax

/generate [options] [-- free-text instruction]

The shell form is rook generate with the same options.

OptionPurpose
--total <n>Approximate target size for the complete suite.
--class <names>Comma-separated classes. Default: functional,adversarial.
--category <names>Comma-separated scenario categories.
--forceRe-derive scenarios even when feature hashes are unchanged.
--allow <rule>Pre-authorize one exact tool rule for this launch. Repeatable.
--jsonEmit machine-readable events.
--verboseShow subagent activity and credits as work happens.
/generate --total 20 --class functional,adversarial
/generate --category boundary,reliability -- emphasize retries and duplicate requests

Generation reads the current feature model, plans coverage, writes scenarios, and checks runnability. It does not invoke the live target. Unchanged features reuse their scenarios without a model call; --force intentionally bypasses that optimization.

Generated files remain editable. A scenario whose origin is human is not silently replaced during later generation.

Scenario guide · /explore · /scenarios · /sync

Test across 3000+ combinations of browsers, real devices & OS.

×
Schedule Your Personal Demo
Book Demo

Help and Support

Related Articles