pushback-templates

Provides phrasing templates and example dialogues for pushing back on user requests.

1|Updated Apr 3, 2026
One-click install
npx skills add https://github.com/TierOne-Studio/spa-velocity --skill pushback-templates-tierone-studio
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pushback-templates
Source: https://github.com/TierOne-Studio/spa-velocity/tree/main/.ruler/skills/pushback-templates
Command: npx skills add https://github.com/TierOne-Studio/spa-velocity --skill pushback-templates-tierone-studio

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When an AI assistant spots a simpler alternative, scope creep, a hidden risk, or a flawed premise in a user request, it often either stays silent or disagrees poorly. This Skill supplies structured templates so pushback is clear, brief, and non-adversarial. ## Core Features & Use Cases - Four trigger-based templates: simpler alternative spotted, scope creep risk, hidden risk, and genuine disagreement with framing, each with recommended phrasing and variants. - Core three-part pattern: state the observation, name the tradeoff, ask the question with a named default. - Rules of engagement: one round of pushback then accept, no moralizing, brevity first, and calibrated confidence. - Use Case: A user asks to render user-supplied markdown with dangerouslySetInnerHTML. The Skill guides the assistant to flag the XSS risk, list mitigations (react-markdown sanitizer, DOMPurify, or accepting the risk), and name a safe default. ## Quick Start Use the pushback-templates skill to help me disagree with this request clearly and briefly, naming the tradeoff and a default option.

Frequently Asked Questions about pushback-templates

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I push back on a user request without sounding adversarial?

Use the three-part pattern: state the observation, name the tradeoff, then ask a question with a named default. Keep it under a few sentences, avoid moralizing, and accept the user's decision after one round of surfacing your concern.

When should an AI assistant disagree with a user's framing?

Disagree when the request rests on a premise contradicted by evidence in code, data, or constraints. Cite the evidence, acknowledge you might be missing context, and propose the better-fitting action before proceeding.

How do I handle scope creep in an implementation request?

Flag that the change would touch areas the user did not name, explain why, and offer two options: ship only the requested piece with a follow-up, or explicitly expand scope. Let the user choose rather than silently expanding.

What are common pushback anti-patterns to avoid?

Avoid gatekeeping, vague fishing questions like 'are you sure?', pushback without a default option, multi-round arguing after the user replies, and pushing back on every request. Surface concerns concretely once, then execute.

When should I not use pushback templates?

Do not use them for normal conversation, agreement situations, or factual questions. The templates apply only when you have spotted a simpler alternative, scope creep, a hidden risk, or a genuinely flawed premise.