Guide question formulation to align with user intent.
npx skills add https://github.com/onblueroses/strata --skill ask-better
Please help me install this Agent Skill. Skill: ask-better Source: https://github.com/onblueroses/strata/tree/main/skills/ask-better Command: npx skills add https://github.com/onblueroses/strata --skill ask-better
This Skill helps align the AI's understanding with the user's intent by guiding the formulation of questions with minimal cognitive load.
Before formulating a question using AskUserQuestion, use the ask-better skill to ensure it aligns with the user's intent and is necessary.
High-intent search queries and answers about installing and using this skill.
AI interaction alignment during code review improves by guiding the formulation of user questions with minimal cognitive load. This ensures the AI's internal model accurately reflects the user's actual intent before executing tasks.
Tiered questioning in AI decision-making prioritizes inquiries based on importance, categorizing them into blocking, shaping, and polishing tiers. This structured approach maximizes alignment efficiency and clarifies user constraints.
A confidence check evaluates the need for a question based on the AI's current understanding of the context. It determines whether asking the user is necessary, reducing unnecessary interruptions and streamlining information gathering.
Use model exposure to enhance AI questioning efficiency when you need the AI to state assumptions and test them transparently. This approach improves clarity and ensures questions align strictly with user goals.
Yes, this questioning approach requires a clear understanding of the user's context and goals beforehand. Establishing this baseline is necessary to effectively minimize the delta between AI understanding and user intent.
IbIFACE-TechAnswer user questions with structured, accurate responses from available context.
arthrodApply meta-cognitive reasoning frameworks to improve AI response quality.
r3dlexGuide teams through planning, review, and decision-making before executing significant tasks.
savvidesAudit course objectives against five adversarial dimensions and output a confidence score.
