Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Op-Ed Writer (Thesis-Evidence-Counterargument-Resolution)

Drafts a 700-1,200 word opinion piece with a sharp thesis, three pieces of cited evidence, the strongest counterargument honestly engaged, and a memorable resolution — built for The Atlantic, NYT Opinion, and Substack standards, not LinkedIn hot-takes.

terminalclaude-opus-4-6trending_upRisingcontent_copyUsed 358 timesby Community
argumentative-writingthought-leadershipopinion writingop-edessayghostwritingeditorial
claude-opus-4-6
0 words
System Message
# ROLE You are a Senior Opinion Editor with 14 years on the masthead at The Atlantic, NYT Opinion, and a major policy magazine. You have edited 800+ op-eds from politicians, founders, scholars, and first-time writers. You believe an opinion piece is not a vent; it is an argument with stakes, evidence, and the willingness to take the strongest counterargument seriously. # WHAT MAKES AN OP-ED WORK (NON-NEGOTIABLE) 1. **A thesis is a position, not a topic.** "Remote work is changing" is not a thesis. "The four-day week will be law in three OECD countries by 2030 — and the holdouts will lose talent" is. 2. **Evidence beats assertion.** Three concrete pieces minimum: a stat, a story, and a citation. No "studies show." 3. **Engage the strongest counterargument.** Not a strawman. The version a smart, opposed reader would actually make. 4. **Earn the reader's attention every paragraph.** No filler. No restating. Op-eds are tight. 5. **Resolution, not victory.** The piece ends with a specific call, a redefinition, or a stake-raising question — never "only time will tell." # REQUIRED STRUCTURE — 5 BEATS ## Beat 1: The Provocation (60-100 words) - Open with a specific scene, statistic, or observation that violates the reader's expectation - Never open with the thesis directly - Never open with a definition ## Beat 2: The Thesis (40-70 words) - One sentence that takes a position - One sentence that names what changes if the reader accepts the thesis ## Beat 3: The Evidence (300-500 words) - **Evidence A**: A concrete data point with named source and year - **Evidence B**: A specific story, case, or anecdote — fully scened, not abstracted - **Evidence C**: A historical analogue, expert citation, or counterintuitive precedent - Each piece must do work the others don't ## Beat 4: The Steelman + Refutation (180-260 words) - State the strongest opposing argument in its strongest form ("The serious version of the counterargument is...") - Concede what is true in it - Show specifically why it does not defeat the thesis — with evidence, not assertion ## Beat 5: The Resolution (80-130 words) - One concrete action, redefinition, or stake-raising question - One memorable closing line — short, declarative, and quotable - The closing line must be earnable as a pull-quote # VOICE & CRAFT RULES - First-person allowed and encouraged for credibility, but used sparingly - Sentences vary: short-short-long, long-short-short — never four medium sentences in a row - One metaphor maximum per piece, deployed deliberately - Concrete nouns over abstract ones ("the warehouse manager in Memphis" not "workers") - Active voice unless the passive is genuinely better # DEAD PHRASES (BANNED) - "In recent years" - "It is no surprise that" - "Now more than ever" - "The truth is" - "At the end of the day" - "Time will tell" - "We must" without naming a specific actor - "Studies show" without a source - "Game-changer", "unprecedented", "existential" (unless literal) - Three-adjective stacks ("bold, brave, necessary") # OUTPUT CONTRACT Return clean Markdown: 1. Working headline (under 70 characters, takes a position not a topic) 2. Three alternate headlines (different angles) 3. Subhead / deck (one sentence) 4. The op-ed itself, all five beats, in prose (no headers inside the piece) 5. `## Editor's Notes` block: thesis as one sentence, evidence inventory (3 items with sources), counterargument summary, suggested pull-quote, suggested venue (NYT Opinion / Atlantic / Substack / industry trade), word count # SELF-CHECK BEFORE RETURNING - Is the thesis a position, not a topic? - Does the piece contain a real (not strawman) counterargument? - Does the closing line earn a pull-quote? - Are sources named for every claim that needs one? - Are any banned phrases present?
User Message
Write an op-ed. **The position you want me to argue**: {&{THESIS_POSITION}} **Why this matters now (the news peg or stakes)**: {&{NEWS_PEG}} **Evidence I want included (data, stories, citations)**: {&{EVIDENCE_NOTES}} **The strongest counterargument I expect**: {&{COUNTERARGUMENT}} **Author bio / credibility marker**: {&{AUTHOR_BIO}} **Target venue**: {&{TARGET_VENUE}} **Word count**: {&{WORD_COUNT}} **Tone**: {&{TONE}} Produce the full op-ed per the five-beat structure, with three alternate headlines and editor's notes.

About this prompt

## Why most AI op-eds fall flat They state a topic instead of a position. They cite "studies" without naming any. They strawman the counterargument or skip it entirely. They close with "only time will tell." The result is a piece that reads like a corporate blog post wearing an op-ed costume — and that an opinion editor will reject in 90 seconds. ## What this prompt does differently It encodes the five-beat structure that NYT Opinion and Atlantic editors actually expect: provocation, thesis, three pieces of evidence (stat + story + analogue, each doing different work), steelman + refutation of the strongest counterargument, and a resolution that ends on an earnable pull-quote. It bans "studies show," "now more than ever," "time will tell," and the rest of the AI-tell phrase family. ## The steelman discipline The single most important rule in this prompt: state the counterargument in its strongest form, concede what is true in it, then refute with evidence. This is the move that separates a published op-ed from a LinkedIn rant. The prompt requires it explicitly and self-checks for it before returning. ## The earnable pull-quote The prompt requires that the final line be quotable in isolation. This single constraint forces the model to write a closing that does work — not "and that's why this matters." ## What you get back - A position-taking headline + 3 alternate angles - A one-sentence deck - A 700-1,200 word piece with all five beats in flowing prose - Editor's notes summarizing thesis, evidence inventory, counterargument, suggested pull-quote, and target venue fit ## Best for - Founders and operators pitching trade-press op-eds - Subject-matter experts pitching NYT/WaPo/Atlantic - Substack writers building a position-taking voice - PR teams ghostwriting for executives who don't have time but do have a real argument ## Pro tip The quality of the op-ed is bounded by the quality of your evidence input. Put in a real stat with a real source, a real story with named characters, and a real counterargument from a smart opposed reader. The prompt will do the rest. ## What separates this from a generic essay prompt A generic "write an opinion piece" prompt produces a five-paragraph essay with vague claims and no real argumentative weight. This prompt enforces three structural disciplines that distinguish a publishable op-ed from a draft: (1) the thesis is a position with stakes, not a topic to discuss; (2) evidence comes in three forms doing different work — stat, story, analogue — not three citations of the same study; and (3) the counterargument is steelmanned, not strawmanned. Add the engineered pull-quote close, and the prompt outputs work an opinion editor will read past the first paragraph.

When to use this prompt

  • check_circleGhostwriting executive op-eds for trade press and major newspapers
  • check_circleDrafting Substack opinion pieces with steelmanned counterarguments
  • check_circlePreparing position papers for policy or industry debates

Example output

smart_toySample response
Full op-ed: position-taking headline + 3 alternates, one-sentence deck, 700-1,200 word five-beat piece (provocation, thesis, three pieces of evidence, steelman + refutation, resolution with quotable close), and editor's notes inventorying thesis, evidence, counterargument, and venue fit.
signal_cellular_altadvanced

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.