bmad-advanced-elicitation

You can systematically refine LLM output by applying iterative elicitation methods like Socratic questioning and pre-mortem analysis. This process enforces critique cycles and structured revisions to surface hidden assumptions and improve response robustness.

The best way to run a red-team review on generated content is using a method registry that applies structured elicitation techniques. It executes ordered critique cycles, allowing you to reshuffle methods and explicitly apply or discard revisions.

Iterative prompt engineering improves multi-agent workflows by enforcing ordered execution and halt conditions during content refinement. This allows persona-driven revisions and section-level drafting to be preserved or discarded only after explicit user confirmation.

Yes, you need a methods.csv registry to load and interpret diverse elicitation techniques like first-principles and red-team reviews. This registry provides the method metadata required to select and execute tailored refinement steps.

Yes, you can reshuffle elicitation methods before applying them to your draft. The interactive refinement process presents up to five candidate methods, executes your selected options in order, and waits for explicit confirmation before overwriting content.

You should avoid automated content refinement when uncontrolled overwrites pose a risk to document production. The elicitation process mitigates this by enforcing explicit apply and confirm steps, but manual review is still recommended for critical sections.