Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for ghostwriters

Nonfiction Outline Evidence Agent

An outline where every chapter is backed by available evidence.

Nonfiction Outline Evidence Agent: what goes in, what the agent does and what you get

What it does

Nonfiction book proposals often promise chapters that the research cannot support, which becomes a problem halfway through writing. This agent maps the available evidence to each chapter's central claim. It picks the weakest claim and runs targeted research for it, using approved databases and the author's source list. If the claim becomes supported, it updates the map and moves to the next weakest. If support stays absent, it proposes narrowing the chapter rather than inventing evidence. It then revises the outline with gaps stated plainly. The author decides on the final outline. Edge case: a claim supported only by anecdotes is not treated as evidence of a trend.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueNo 1 STARTS WHEN Proposal drafted 2 USES A TOOL Map evidence to chapter claims 3 DOES Pick the weakest claim 4 USES A TOOL Run targeted research 5 CHECKS THE RESULT Is the claim now supported? If not: propose narrowing the chapter. Back to step 3. 6 DOES Revise the outline with gaps 7 RESULT Evidence-aware outline
Read the steps as a list
  1. Proposal drafted
  2. Map evidence to chapter claims
  3. Pick the weakest claim
  4. Run targeted research
  5. Is the claim now supported?If not: propose narrowing the chapter. Back to step 3.
  6. Revise the outline with gaps
  7. Evidence-aware outline

How it decides

It researches the weakest central claim first.

  • Never invent evidence for a claim.
  • Anecdotes do not prove a trend.
  • Narrow the chapter when support stays absent.
  • Work on the weakest claim first.

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Approved research databases
  • Research budget per claim (default 2 hours)
  • How many weak claims to work on per run (default 3)
  • Who approves the outline (default: the author)
  • Outline format

What keeps you in control

It always asks you first

  • Publication
  • Final factual conclusions

Hard limits

  • No invented evidence.

It stops when

  • Done: outline revised.

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA proposal for a book on remote work has 12 chapters. The agent finds chapter 7, on a 'trend' of rural tech hubs, supported only by 3 anecdotes. Targeted research finds no data showing a trend, so the check fails. It proposes narrowing the chapter to 3 case studies. The author agrees and approves the revised outline before sending it to her agent.

More agents for ghostwriters