Latest Claude Code command - BTW #aitools #claudecode #anthropic

DevNotes · Intermediate ·🧠 Large Language Models ·3mo ago

Key Takeaways

The video demonstrates the use of the /btw command in Claude Code, a feature that allows users to ask side questions without interrupting the main conversation or messing up the context. The command enables Claude to answer questions on the side, reusing the prompt cache and minimizing the use of extra tokens.

Full Transcript

So Claude's in the middle of refactoring your code and you want to ask it something like hey what pattern is this file using or wait what does this function do but you don't want to interrupt what it's doing or mess up the conversation that's /btbu you just type /btw and then your question and claude answers it on the side without touching your main conversation it sees everything in your context but it doesn't add anything to it so your session stays clean and the best part is it basically reuses your prompt cache so it's almost free. Like you could ask 10 side questions during a long refactor and barely use any extra tokens. Think of it as the inverse of a sub aent. A sub aent gets tools but empty context/BTW gets your full context but no tools. Follow for more Claude Code tips.
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

The /btw command in Claude Code allows users to ask side questions without interrupting the main conversation. This feature reuses the prompt cache, minimizing extra token usage. The command is useful for refactoring code and managing context.

Key Takeaways
  1. Type /btw followed by your question to ask a side question
  2. Claude will answer the question on the side without touching the main conversation
  3. The command reuses the prompt cache, minimizing extra token usage
  4. Use the /btw command to ask multiple side questions during a long refactor
  5. The command is useful for managing context and avoiding interruptions
💡 The /btw command is the inverse of a sub aent, providing full context but no tools, allowing users to ask side questions without interrupting the main conversation.

Related Reads

📰
GPT-5.6 USA: Global Privacy & Competition Ripple
Understand the global implications of the US GPT-5.6 rollout on AI regulation and competition
Dev.to AI
 📰
Integrating the OpenAI API the Right Way — Streaming, Rate-Limiting, and Prompt
Learn to integrate the OpenAI API effectively with streaming, rate-limiting, and prompt engineering for better performance and results
Dev.to · Ardhansu Das
 📰
Building an NLP Pipeline That Actually Understands Offer Text (with spaCy)
Learn to build an NLP pipeline using spaCy that extracts structured data from offer text, improving understanding and analysis
Dev.to · Ardhansu Das
 📰
Building an AI Web Crawler That Outputs LLM-Ready Content Chunks
Learn to build an AI web crawler that outputs LLM-ready content chunks, enabling efficient data preparation for large language models
Dev.to · Oaida Adrian
Up next
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)
Watch →