Free prompt library

Prompts that actually do real work.

32 battle-tested prompts for GEO audits, AI search visibility, content optimization, and AI workflow design. Copy any one and run it today. No signup required.

Filter by tag:

GEO Audit

7 prompts
Intermediate Perplexity Sonar Pro Search, Sonar Deep Research, or ChatGPT Deep Research (requires live web access)

AI Overview Presence Tracker

Determine whether a target query triggers a Google AI Overview, who is cited inside it, and what content shape the cited sources share.

Best for: Track AI Overview presence on priority commercial queries
View prompt →
Beginner Any frontier model: Claude Opus 4.7, GPT-5.5, Gemini 3.1 Pro, or Perplexity Sonar Pro

Brand-in-Answer Visibility Check

Quick single-prompt check to see whether a target brand appears in an AI-generated answer for a specific buyer query.

Best for: Quick spot-check during a launch, campaign review, or before a sales call
View prompt →
Beginner Perplexity Sonar Pro Search or ChatGPT Deep Research (requires live web access)

Branded SERP Audit for Entity Authority

Audit what surfaces when someone searches your brand name: knowledge panel, AI Overview, social profiles, reviews, and any low-quality results that can erode trust.

Best for: Quarterly entity health check and post-launch reputation review
View prompt →
Intermediate Claude Opus 4.7, GPT-5.5, or Gemini 3.1 Pro (run identical prompt across each engine)

Citation Share Audit Across AI Engines

Run a fixed panel of high-intent prompts across ChatGPT, Claude, Perplexity, and Google AI Overviews to score brand visibility and citation share.

Best for: Run weekly to track brand visibility across AI search engines
View prompt →
Intermediate Claude Opus 4.7 or GPT-5.5

Competitor Citation Comparison

Score your brand against three competitors on the same prompt panel to identify where you are winning, losing, or invisible in AI search.

Best for: Quarterly competitive review, board reporting, content gap planning
View prompt →
Intermediate Claude Opus 4.7 or GPT-5.5

E-E-A-T Signal Audit for a Page

Score a page's expertise, experience, authoritativeness, and trust signals using the same dimensions Google's quality raters and AI retrieval systems weight.

Best for: Pre-publish QA on pillar pages and service pages
View prompt →
Intermediate Perplexity Sonar Deep Research or ChatGPT Deep Research (live web required)

Reddit and Forum Citation Finder

Find every Reddit thread, Stack Exchange answer, Hacker News post, and niche forum discussion where your brand or category is mentioned, scored by AI-citation potential.

Best for: Quarterly forum-presence audit and content seeding planning
View prompt →

Content & Editorial

10 prompts
Beginner Claude Opus 4.7 or GPT-5.5

Headline Variant Tester

Generate 5 headline variants for any piece of content and rank them on three dimensions (specificity, curiosity, fit-for-buyer) with rationale per variant.

Best for: Pre-publish headline review for blog posts, email campaigns, and ads
View prompt →
Beginner Claude Opus 4.7 or GPT-5.5

Newsletter Draft from Raw Notes

Take 5 to 15 bullets of rough notes and draft a coherent newsletter issue with a hook, one main idea, and a clean closing.

Best for: Drafting weekly or monthly newsletter issues from collected notes
View prompt →
Intermediate Claude Opus 4.7, GPT-5.5, or Gemini 3.1 Pro

Comparison Page Outline Generator

Generate the structure for a 'X vs Y' or 'best X' comparison page that ranks in Google, gets cited in AI Overviews, and converts buyers in a comparison stage.

Best for: Planning category comparison pages and 'best of' guides
View prompt →
Intermediate Claude Opus 4.7 (best for nuanced rewrites) or GPT-5.5

Entity Clarity Rewrite for Service and About Pages

Tighten entity references on a key page so AI engines can reliably identify what the brand is, who it serves, and how it fits in its category.

Best for: Auditing and rewriting brand pages for entity clarity
View prompt →
Beginner Claude Opus 4.7, GPT-5.5, or Gemini 3.1 Pro

FAQ Generator from an Existing Page

Extract real buyer questions from an existing page and generate FAQ-schema-ready answers that match the page's voice and substance.

Best for: Adding FAQPage schema to service pages and pillar content
View prompt →
Beginner Claude Opus 4.7 or GPT-5.5

Glossary Entry Generator for AI Citation

Write clean, definition-style glossary entries optimized for AI engines to extract and cite directly. Each entry follows the structure retrieval systems prefer.

Best for: Building category glossaries, definition pages, and AI-citation-friendly reference content
View prompt →
Beginner Claude Opus 4.7 or GPT-5.5

Internal Linking Audit for a Page

Identify missing internal links from a page to other relevant pages on the site, with specific anchor-text suggestions and a priority order.

Best for: Tightening a topic cluster after publishing new content
View prompt →
Beginner Claude Opus 4.7 (best for nuanced rewrites) or GPT-5.5

Passage Extraction Rewrite for AI Citation

Rewrite existing paragraphs into self-contained, extractable passages that AI engines can pull and cite directly in generated answers.

Best for: Refactoring existing pages for AI search visibility
View prompt →
Intermediate Claude Opus 4.7 or GPT-5.5 (use structured-output mode if available)

Schema Markup Generator from a Page

Generate validated JSON-LD schema markup (Article, Product, Service, FAQ, HowTo, Organization, BreadcrumbList) for any page based on its actual visible content.

Best for: Adding or fixing structured data on priority pages
View prompt →
Intermediate Claude Opus 4.7 or GPT-5.5

Topic Cluster Builder

Design a pillar-and-spoke content architecture for a topic, including the pillar page, supporting articles, internal link map, and priority order.

Best for: Quarterly content roadmap planning and topical authority audits
View prompt →
Intermediate GPT-5.5 or Claude Opus 4.7

Hiring Rubric Generator

From a job description, produce a structured candidate rubric with skills weights, evaluation prompts per skill, and red flags to listen for during interviews.

Best for: Standardizing candidate evaluation before opening a hiring loop
View prompt →
Intermediate GPT-5.5 or Claude Opus 4.7

Vendor Selection Scorecard

Compare 3 to 5 vendors against named criteria with weighted scores, plus a 'watch list' of risks per vendor and a single recommendation.

Best for: Software procurement, agency selection, or partner evaluation
View prompt →
Beginner GPT-5.5 or Claude Opus 4.7

AI Readiness Assessment for a Business Process

Evaluate which parts of an existing business process are agent-ready, which need human-in-the-loop, and which should stay fully manual.

Best for: Pre-deployment audit before introducing AI agents to a workflow
View prompt →
Intermediate GPT-5.5 or Claude Opus 4.7

Content Refresh Prioritization

Score a backlog of older posts by refresh ROI and decide which to update first, which to consolidate, and which to retire.

Best for: Quarterly editorial planning and content backlog triage
View prompt →
Intermediate GPT-5.5 or Claude Opus 4.7

Human Handoff Map for Agent Workflows

Generate a documented list of every point in an agent-driven workflow where a human must review or approve before the next step runs.

Best for: Scoping the boundaries of agent autonomy before deployment
View prompt →
Intermediate Claude Opus 4.7 or GPT-5.5

Meta-Prompt Optimizer

Take a weak or inconsistent prompt and rewrite it using 2026 best practices: XML structure, literal instructions, output contracts, and hallucination guards.

Best for: Improving any prompt that produces inconsistent output
View prompt →

Productivity

5 prompts
Beginner Claude Opus 4.7 or GPT-5.5

Calendar Audit for Focus Time

Take a week of meeting titles and durations and return a focus-time audit identifying meetings to decline, batch, shorten, or replace with async.

Best for: Quarterly calendar audit and weekly meeting triage
View prompt →
Beginner Claude Opus 4.7 or GPT-5.5

Email Reply Drafter in Three Variants

Draft three reply variants (acknowledge-only, action-oriented, escalating) for an inbound email with specific tone and length controls.

Best for: Processing inbound email faster without sounding like a template
View prompt →
Beginner Claude Opus 4.7 (long context) or Gemini 3.1 Pro (2M-token context window)

Customer Review Synthesis

Turn 50 to 200 customer reviews into a structured insight brief: top themes, sentiment-weighted strengths and weaknesses, persona signals, and content opportunities.

Best for: Quarterly voice-of-customer analysis and content opportunity discovery
View prompt →
Beginner Claude Opus 4.7 (best at attribution) or GPT-5.5

Meeting Notes to Action Items

Convert raw meeting transcripts into clean, owner-tagged action items with due dates that go straight into your task system.

Best for: Post-meeting follow-up
View prompt →
Beginner Claude Opus 4.7 (long context) or Gemini 3.1 Pro (2M-token context window)

Research Synthesis from Multiple Sources

Turn five to ten sources on the same topic into a structured research brief with citations, agreements, disagreements, and a confidence read.

Best for: Pre-meeting prep, market research, competitive intelligence
View prompt →

Download all 32 prompts as a PDF.

Includes every prompt on this page, organized by category, with copy-ready formatting and Arthur's notes on when to use each one. Free. One email, one download.

No spam. Unsubscribe anytime. Your email is never shared.

Need a custom prompt library for your team?

I help teams design, test, and document prompt libraries that hold up under production use. Reusable, evaluated, and integrated into the workflows that matter.