liveBook · Manning

Dialogue

As explained in Claude at Work

A conversation is a cumulative interaction session consisting of messages exchanged between a user and an AI assistant.

A conversation is a cumulative session: every prompt and response adds to its context. Tool use, such as searching the web and loading pages, can add thousands of tokens in a turn. section 1.4.1

Each new chat starts with a clean slate. Supply relevant context rather than assuming it carries over; manage sessions by renaming chats, favoriting them, returning to old ones, requesting summaries, or exporting transcripts. section 1.4.1 section 2.9

Use a deliberate workflow to keep the exchange useful:

  • Open with a goal, context, constraints, and desired format. section 2.9
  • Evaluate responses, iterate on gaps, and push back with evidence when something is wrong. section 2.9
  • If a thread becomes confused, start a fresh session with a stronger opening prompt. section 2.9

Opening prompt template

Goal: [task]. Context: [relevant facts]. Constraints: [limits]. Format: [desired output]. section 2.9

Figure 1.2 How context builds over the turns of a conversation. The input grows cumulatively and there can be huge jumps from tool use when Claude is researching file content or web pages.
CH01 F03 Stoneman

The diagram shows context accumulating across turns, with large jumps when tools load research or file content. section 1.4.1

How Claude at Work covers it

How deep it goes
Explains cumulative token use and practical session management, but gives no detailed context-compression procedure.
Examples
Illustrates context growth through turns and tool-driven research, and recommends prompts structured around goals, context, constraints, and format.
What it assumes
Assumes users will supply relevant context such as date, location, or tech stack, and teaches prompt structure and session management directly.
What you can do after
Operate Claude sessions by structuring openings, revisiting or summarizing chats, and restarting confused threads.
How it connects
Connects prompt structure, token limits, context windows, tool use, artifacts, and output evaluation.

Also explained in

Compare how 12 books explain dialogue →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Claude at Work. The section links point to the parts of the book it draws on.