smart-followups_skill

Overview

This skill generates three contextual follow-up suggestions after any AI response when the user types /followups (or an alias). It presents Quick, Deep Dive, and Related options as clickable buttons in supported channels or as numbered text choices in plain-text channels. The goal is to surface immediate clarifications, technical depth, and related topics to keep conversations productive and focused.

How this skill works

When triggered, the handler captures recent conversation context and prompts the current model to produce three targeted follow-up prompts: a Quick clarification or next step, a Deep Dive for technical detail, and a Related topic for broader context. Output is formatted as interactive buttons in channels that support them, or as numbered text suggestions where buttons are unavailable. Selecting or replying with a suggestion sends that follow-up back to the model to continue the discussion.

When to use it

• After a high-level answer to get a clear next step or clarification
• When you want a deeper technical explanation without crafting a new prompt
• To explore related topics or alternatives relevant to the current thread
• On messaging platforms with buttons to speed up conversational flow
• In CLI or text-only channels where choosing 1–3 is faster than typing

Best practices

• Trigger /followups immediately after an answer to keep context fresh
• Use Quick for immediate actions or simple clarifications
• Pick Deep Dive when you need step-by-step guidance or technical rationale
• Choose Related to discover alternatives, comparisons, or broader implications
• If suggestions aren’t relevant, reply with a custom prompt referencing the prior reply

Example use cases

• You ask what Docker is, then tap Quick to get installation steps, Deep Dive for container architecture, or Related for orchestration options
• After a legal or compliance summary, request Deep Dive to see clause-level explanations
• In a coding help thread, use Quick to get a minimal reproducible example and Deep Dive to optimize performance
• When evaluating tools, use Related to compare competitors or integration options
• In teaching scenarios, generate follow-ups to scaffold student understanding progressively