Build state

• /_books/hello-claude.md updated with current-documented platform behavior and conservative, evergreen wording (2026-05-30).
• Upstream blocker for regenerated PDF/EPUB has been cleared; SHA-2275 is now an editorial proposal/ownership handoff rather than an artifact-generation blocker.
• Current ownership note lives at memory/hello-claude-ownership-note-2026-05-31.md.

Proposed title

Field Guide 2: Asking Better Questions with Claude

Review ask for Avery

Approve, rename, or redirect this next-guide concept. My recommendation is to proceed because it naturally follows Hello, Claude: the first guide teaches comfort and safety; the second teaches repeatable prompting habits without drifting into technical prompt-engineering jargon.

Intended audience

Readers who have opened Claude, tried a few everyday requests, and now want more consistent results without becoming technical prompt engineers. They understand the basics from Hello, Claude, but still run into vague answers, overlong answers, weak drafts, or factual claims they are not sure how to check.

One-paragraph promise

This guide is for readers who have used Claude a little but still get mixed results because they don’t know how to ask for complex work without overloading or under-specifying. It teaches a practical question-design method for higher-stakes tasks—research, planning, writing, analysis, and decision support—so readers can get better outputs with fewer edits. It emphasizes trust calibration, source verification, and output shaping so readers can work with Claude in a way that is faster, safer, and more repeatable at home, at work, and for side projects.

Proposed chapter outline

1. Why Claude quality follows your question quality
   • The simple loop: intent → prompt → context → constraints → refinement.
   • What makes one answer useful vs noisy or vague.
2. The two anchors: goal and success criteria
   • Define success before prompting.
   • Convert fuzzy goals into measurable results.
3. Context without clutter
   • What to include, what to leave out.
   • How to provide source and tone constraints clearly.
4. Output contracts
   • Asking for structure: tables, checklists, first drafts, summaries, comparatives.
   • How format requests reduce rework.
5. Iterative improvement workflow
   • A practical loop: draft, critique, constrain, refine.
   • Common failure patterns and quick rewrites.
6. Working with external material
   • Documents and images: what can be uploaded and what to expect.
   • How to stage projects for long-running tasks.
7. Fact discipline
   • How to force confidence checks.
   • How to prompt for uncertainty, caveats, and verification sources.
8. Decision support, not decision replacement
   • Framing trade-offs, risk, and assumptions with Claude.
   • What to verify before acting.
9. Templates you can reuse
   • 10 reusable prompt templates for common work: planning, analysis, writing, feedback loops.
10. Your Claude habit system (7 days)
    • A practical ramp-up routine tied to daily use cases.

Source outline file

The working source outline now lives at /_book-src/asking-better-questions-with-claude/outline.md so drafting can begin cleanly after Avery approves or redirects the concept.

Source and verification plan

• Use Anthropic Help Center and Anthropic docs as primary sources for Claude feature behavior, especially Projects, document/image handling, personalization, privacy, plan availability, and any account-specific feature notes.
• Avoid hard-coded pricing, message limits, model availability, token/context numbers, or plan entitlements unless they are checked during the drafting pass and phrased as current-at-publication rather than permanent.
• Keep examples tested in Claude during drafting where possible: email rewrite, document-check, decision-support, source-check, style-match, and refinement loop examples.
• Include a Sources & further reading appendix with primary links, including:
  • https://support.anthropic.com/en/articles/9517075-what-are-projects
  • https://support.anthropic.com/en/articles/9519177-how-can-i-create-and-manage-projects
  • https://support.anthropic.com/en/articles/10185728-understanding-claude-s-personalization-features
  • https://privacy.anthropic.com/en/articles/10023580-is-my-data-used-for-model-training