btw

Run parallel side conversations with an LLM agent during long-running tasks.

12|Updated Feb 2, 2026
One-click install
npx skills add https://github.com/pratos/clanker-setup --skill btw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: btw
Source: https://github.com/pratos/clanker-setup/tree/main/skills/btw
Command: npx skills add https://github.com/pratos/clanker-setup --skill btw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill allows users to have side conversations with the LLM without interrupting the main agent's ongoing tasks, enabling quick questions, planning, or clarifications.

Core Features & Use Cases

  • Parallel Conversations: Engage in separate discussions with the LLM while the main agent is busy.
  • Contextual Awareness: The side conversation can access main session messages and previous side-thread exchanges.
  • Seamless Integration: Responses stream into a widget above the editor, and can be injected back into the main session.
  • Use Case: While the agent is refactoring code, you can use /btw to ask clarifying questions about the changes or plan the next steps for testing.

Quick Start

Ask a quick question in the side conversation by typing /btw what is the current task.

Frequently Asked Questions about btw

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

FAQPage Schema
How do I ask an LLM clarifying questions without interrupting a long-running agent task?

You can use parallel LLM conversations to ask clarifying questions without interrupting a long-running agent task. This allows you to run side chats for planning or quick queries while the primary agent continues its workflow seamlessly.

Can I inject context from a side conversation back into my main agent session?

Yes, you can inject context from a side conversation back into your main agent session. The side conversation supports multi-turn threads and allows you to summarize side conversation exchanges directly into the primary agent's context.

What is the best way to plan next steps while an LLM agent is refactoring code?

The best way to plan next steps while an LLM agent is refactoring code is to start a parallel conversation. You can use side chats to query the LLM about changes or plan testing steps without disrupting the active code refactoring process.

Does a parallel side chat have access to previous messages in the main session?

Yes, a parallel side chat has contextual awareness and can access previous messages in the main session. This ensures your side conversation maintains context from both the ongoing primary session and previous side-thread exchanges.

How do I start a non-disruptive side conversation with an LLM agent?

To start a non-disruptive side conversation with an LLM agent, type your quick question using the side chat command. Responses will stream into a widget above the editor, allowing you to interact without pausing the primary workflow.