askme

To reduce planning ambiguity, you can execute Verbalized Sampling to generate and weight multiple intent hypotheses, then use an AskUserQuestion protocol to surface precise clarifying questions before finalizing your plan.

Verbalized Sampling is a decision-support mechanism that generates and weights multiple intent hypotheses across ambiguous tasks to identify user intent and reduce planning ambiguity before committing to a plan.

You clarify plans by running structured exploration with mode auto-detection across exhaustive, collaborative, and adversarial modes, applying per-hypothesis falsifiers and maximum-question batching to isolate intent.

The best way to surface precise clarifying questions is using an AskUserQuestion protocol, which batches maximum-question queries to guide users through structured exploration and resolve multi-hypothesis ambiguity.

Yes, frontmatter-driven configuration loads directives directly into context to activate mode auto-detection, enabling the system to switch between exhaustive, collaborative, and adversarial exploration modes.

Use adversarial mode to apply per-hypothesis falsifiers that aggressively challenge assumptions, whereas collaborative mode guides users through structured exploration to cooperatively reduce planning ambiguity.