30 Claude Prompts That Write Knowledge Base Articles
Describe the topic and Claude returns a complete, ready-to-publish help-center article as a structured artifact you can paste straight into Zendesk, Intercom, HelpScout, or Notion. Prompts for how-tos, troubleshooting, glossaries, policies, FAQs, and feature guides. Not "give me some notes."
In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 15 prompts are free instantly โ no signup needed. Hand-curated and tested by the AI Academy team.
How-To Articles
5 promptsStep-by-Step Task Walkthrough
1/30You are a senior technical writer who produces clear, scannable help-center articles. <context> I need a complete, ready-to-publish how-to article that walks a user through one specific task, returned as a single self-contained Markdown document I can paste into my help center. </context> <inputs> - Product and what it does: [PRODUCT ONE-LINER] - The exact task the article covers: [E.G. HOW TO INVITE A TEAMMATE] - Who is doing it and their access level: [USER ROLE] - Where in the product it happens: [MENU / SCREEN PATH] - Prerequisites: [WHAT MUST BE TRUE FIRST] - Common mistakes to preempt: [PITFALLS] </inputs> <task> Write the article with: an H1 title phrased as the user's goal, a one-sentence summary of the outcome, a short "Before you start" prerequisites list, numbered steps (each a single action, with the exact button or menu label in bold and a placeholder callout for a screenshot), an expected result after the last step, a "Troubleshooting" mini-section for the 2-3 common mistakes, and a "Related articles" list of 3 linked titles. </task> <constraints> - Valid, self-contained Markdown; one action per numbered step; plain language at roughly an 8th-grade reading level. - Use exact UI labels in bold; no filler, no marketing tone; keep steps testable. - Add a short meta block at the top (title, description under 160 chars, keywords) as an HTML comment for CMS import. </constraints> <format> Return the full article as a Markdown artifact, then a one-line note on where to insert screenshots and how to adapt it for a different user role. </format>
Produces a complete step-by-step how-to article with prerequisites, numbered steps, and troubleshooting, ready to use.
Pro tip: List every exact button label from your UI in the inputs so Claude bolds the real labels instead of guessing generic ones.
Getting-Started / Onboarding Guide
2/30You are a customer-education specialist who writes onboarding guides that get new users to first value fast. <context> I need a ready-to-publish getting-started article for brand-new users, returned as one self-contained Markdown document for my help center. </context> <inputs> - Product and core value: [WHAT IT DELIVERS] - The "aha moment" a new user should reach: [FIRST WIN] - The 4-6 setup steps to get there: [STEP LIST] - Account tiers or plans that change the flow: [IF ANY] - Time it should take: [E.G. UNDER 10 MINUTES] - Where to go for help: [SUPPORT LINK OR CHANNEL] </inputs> <task> Write the guide with: an H1 welcome title, a short paragraph framing what they will achieve and how long it takes, a "What you will need" checklist, a numbered setup sequence grouped into logical phases with H2 headers, a highlighted "You did it" milestone callout at the aha moment, a "Next steps" list linking 3-4 follow-on articles, and a "Need help?" footer. </task> <constraints> - Valid, self-contained Markdown; encouraging but concise tone; each step ends in a verifiable outcome. - Bold the exact UI actions; use a checklist for prerequisites; no jargon without a one-line definition. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full guide as a Markdown artifact, then a short note on how to branch the flow for a different plan tier. </format>
Generates a new-user getting-started guide that leads to the first win, ready to publish.
Pro tip: Tell Claude the single 'aha moment' metric you track and it will structure every step to drive users toward that one action.
Integration / Connect-a-Tool How-To
3/30You are a technical writer who documents third-party integrations for a SaaS help center. <context> I need a ready-to-publish article explaining how to connect my product to a third-party tool, returned as one self-contained Markdown document. </context> <inputs> - My product: [NAME] - The tool being connected: [E.G. SLACK / HUBSPOT / GOOGLE DRIVE] - Connection method: [OAUTH / API KEY / WEBHOOK] - What data flows and in which direction: [DATA + DIRECTION] - Permissions or scopes required: [SCOPES] - Known limits or gotchas: [RATE LIMITS, PLAN REQUIREMENTS] </inputs> <task> Write the article with: an H1 title naming both tools, a summary of what the integration enables, a "Requirements" list (plan level, admin permissions, accounts needed), numbered connection steps split into "In [my product]" and "In [the tool]" sub-sections, a verification step to confirm it worked, a data-and-permissions explainer table, a troubleshooting section for auth and sync failures, and a "How to disconnect" section. </task> <constraints> - Valid, self-contained Markdown with at least one Markdown table (data flow or permissions). - Exact UI labels in bold; call out anything requiring admin access; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on which step most often causes support tickets and how to preempt it. </format>
Builds an integration how-to with requirements, dual-side steps, and a data-flow table, ready to publish.
Pro tip: Specify the exact OAuth scopes or API permissions; Claude will surface them in a table so users know what access they are granting.
Data Import / Export How-To
4/30You are a support-documentation writer covering data migration and export workflows. <context> I need a ready-to-publish article explaining how to import or export data in my product, returned as one self-contained Markdown document. </context> <inputs> - Product: [NAME] - Direction this article covers: [IMPORT / EXPORT / BOTH] - Supported file formats: [E.G. CSV, XLSX, JSON] - Required columns or fields and their formats: [FIELD SPEC] - File-size or row limits: [LIMITS] - What happens to duplicates or errors: [DEDUPE / ERROR HANDLING] </inputs> <task> Write the article with: an H1 title, a summary of what can be moved, a "Before you start" prep list, a field-mapping reference table (column name, required?, format, example value), numbered steps to run the import or export, an explanation of how errors and duplicates are handled, a downloadable-template callout, and a troubleshooting section for the most common failed-row causes. </task> <constraints> - Valid, self-contained Markdown including a field-mapping table with an example row. - Show a small sample of correctly formatted data in a fenced code block; exact UI labels in bold. - No filler; include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to extend the field table when new columns are added. </format>
Creates a data import/export article with a field-mapping table and sample file, ready to publish.
Pro tip: Paste one real header row from your CSV template so Claude builds the field table and sample against your actual schema.
Mobile App How-To Article
5/30You are a technical writer specializing in concise mobile-app help content. <context> I need a ready-to-publish how-to article for doing one task inside my mobile app, returned as one self-contained Markdown document. </context> <inputs> - App name and platforms: [NAME, IOS / ANDROID / BOTH] - The task the article covers: [E.G. ENABLE NOTIFICATIONS] - Navigation path in the app: [TAB / ICON / MENU SEQUENCE] - Any platform differences: [IOS VS ANDROID DIFFERENCES] - Permissions the OS will prompt for: [CAMERA, LOCATION, ETC] - Version requirement: [MIN APP OR OS VERSION] </inputs> <task> Write the article with: an H1 goal-phrased title, a one-line outcome summary, a version and platform note, numbered steps using gesture-accurate language (tap, swipe, long-press) with the exact icon or label in bold, separate short call-outs where iOS and Android differ, a note on any OS permission prompt and why it is needed, and a "Still not working?" troubleshooting list of 3 fixes. </task> <constraints> - Valid, self-contained Markdown; mobile-accurate verbs; steps short enough to follow on a phone. - Clearly flag any step that differs by platform; no desktop-only assumptions; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to split it into two versions if the iOS and Android flows diverge too much. </format>
Produces a mobile-specific how-to article with gesture-accurate steps and platform notes, ready to publish.
Pro tip: Tell Claude where iOS and Android diverge and it will add inline platform call-outs instead of writing two whole articles.
XML tags are just the start. Learn the full Claude workflow.
A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.
Troubleshooting Articles
5 prompts"Why Isn't X Working" Troubleshooter
6/30You are a support engineer who writes calm, systematic troubleshooting articles. <context> I need a ready-to-publish troubleshooting article for a single symptom users report, returned as one self-contained Markdown document. </context> <inputs> - Product: [NAME] - The symptom as users describe it: [E.G. MY EXPORT IS STUCK] - The most likely causes, most common first: [CAUSE LIST] - Quick checks the user can do: [SELF-SERVE CHECKS] - When to escalate to support: [ESCALATION TRIGGER] - Info support will need if escalated: [DIAGNOSTIC DETAILS] </inputs> <task> Write the article with: an H1 phrased as the symptom, a reassuring one-line intro, a "Quick fixes to try first" ordered list, then a cause-by-cause section where each cause has a short "How to check" and "How to fix" pair, a clear "If none of this works" escalation block listing exactly what to send support, and a "Related issues" list of 3 linked articles. </task> <constraints> - Valid, self-contained Markdown ordered from cheapest/most-common fix to rarest. - Each fix is a concrete action, not "contact us"; exact UI labels in bold; empathetic but efficient tone. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to reorder the causes if your ticket data shows a different top offender. </format>
Generates a symptom-based troubleshooting article ordered from common to rare fixes, ready to publish.
Pro tip: Rank the causes by how often they appear in your tickets; Claude will keep that order so users hit the likely fix first.
Error-Code Reference Article
7/30You are a documentation engineer who writes error-message reference articles. <context> I need a ready-to-publish reference article that lists error codes or messages and how to resolve each, returned as one self-contained Markdown document. </context> <inputs> - Product or API: [NAME] - The errors to document (code, message, when it fires): [ERROR LIST] - Category grouping if any: [AUTH / VALIDATION / RATE-LIMIT] - The audience: [END USERS / DEVELOPERS] - Where the error appears: [UI BANNER / API RESPONSE / LOG] - Support contact for unresolved cases: [CHANNEL] </inputs> <task> Write the article with: an H1 title, a short intro on how to read the reference, a searchable reference table with columns for Code, Message, What it means, Likely cause, and Fix, then an expanded H3 sub-section for the 3-5 most common errors each with a plain-English explanation, a fix checklist, and a sample of the raw error in a fenced code block, and a footer on how to report an undocumented error. </task> <constraints> - Valid, self-contained Markdown with one reference table plus expanded sections. - Each Fix must be an actionable step; show raw error text in code blocks; no vague advice. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to keep the table alphabetized and add new codes without breaking the structure. </format>
Builds an error-code reference with a lookup table and expanded fixes for common errors, ready to publish.
Pro tip: Give Claude the raw error strings verbatim; it will place them in code blocks so users can Ctrl+F the exact message.
Login & Access Troubleshooting
8/30You are a support specialist who writes account-access help articles. <context> I need a ready-to-publish troubleshooting article for users who cannot log in or access their account, returned as one self-contained Markdown document. </context> <inputs> - Product: [NAME] - Auth methods supported: [PASSWORD / SSO / MAGIC LINK / 2FA] - Common lockout causes: [WRONG PASSWORD, EXPIRED SESSION, ETC] - Password-reset and 2FA-recovery paths: [HOW THEY WORK] - What only an admin can fix: [ADMIN-ONLY ISSUES] - Security note to include: [E.G. WE NEVER ASK FOR YOUR PASSWORD] </inputs> <task> Write the article with: an H1 title, a one-line reassurance, a "Start here" quick checklist (correct email, caps lock, connection), then sections by scenario: forgot password, 2FA device lost, SSO error, account locked, and email-not-recognized, each with concrete steps and a recovery path; an admin-vs-user responsibility note; a security reminder about never sharing credentials; and an escalation block for identity verification. </task> <constraints> - Valid, self-contained Markdown organized by scenario with clear headers users can scan to. - Never instruct users to share passwords or codes; exact UI labels in bold; calm, secure tone. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on which scenario to feature first based on your auth setup. </format>
Creates a login and access troubleshooting article covering resets, 2FA, and SSO by scenario, ready to publish.
Pro tip: Tell Claude exactly which recovery paths are self-serve vs admin-only so it never sends users down a dead-end they can't complete.
Performance / Slowness Diagnostic
9/30You are a support engineer who writes performance-troubleshooting articles. <context> I need a ready-to-publish article helping users diagnose slowness, lag, or timeouts in my product, returned as one self-contained Markdown document. </context> <inputs> - Product: [NAME] - The performance symptom: [SLOW LOADING / TIMEOUTS / LAGGY UI] - Client-side factors that cause it: [BROWSER, EXTENSIONS, CONNECTION] - Product-side or data-size factors: [LARGE DATASETS, HEAVY VIEWS] - Known status-page or maintenance situations: [WHERE TO CHECK] - Diagnostics support may request: [CONSOLE LOG, HAR FILE] </inputs> <task> Write the article with: an H1 title, an intro separating "is it me or the service" (link to status page check), a "Fast checks" list (refresh, clear cache, try another browser, test connection), a section on data-size or heavy-view causes with how to lighten them, a browser and extension conflict section, a step for capturing diagnostics to send support, and an escalation block. Include a short table mapping symptom to first thing to check. </task> <constraints> - Valid, self-contained Markdown with a symptom-to-check table. - Every step is actionable by a non-technical user; explain how to capture a console log or HAR file simply. - No filler; include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to add product-specific diagnostics your team actually needs. </format>
Generates a performance-troubleshooting article with a symptom-to-check table and diagnostic capture steps, ready to publish.
Pro tip: Have Claude include exact steps to capture a HAR file per browser so support gets usable diagnostics on the first reply.
Decision-Tree Troubleshooter
10/30You are a support-content designer who builds branching troubleshooting guides. <context> I need a ready-to-publish decision-tree troubleshooting article that routes users to the right fix based on their answers, returned as one self-contained Markdown document. </context> <inputs> - Product: [NAME] - The broad problem area: [E.G. NOTIFICATIONS NOT ARRIVING] - The key branching questions: [YES/NO OR MULTIPLE-CHOICE QUESTIONS] - The distinct end-fixes each branch leads to: [FIX LIST] - When a branch should escalate: [ESCALATION POINTS] - Anchor links available in my CMS: [YES / NO] </inputs> <task> Write the article as a guided decision tree: an H1 title, a one-line intro on how to use it, a "Start here" question, then a set of clearly labeled branch sections (using headings like "If yes, go to A") that a reader can jump between via anchor links, each terminating in a concrete fix or an escalation step. Also include a compact Mermaid flowchart of the whole tree so authors can see the logic at a glance. </task> <constraints> - Valid, self-contained Markdown; branches use in-page anchor links so users can navigate without scrolling blindly. - Include one fenced Mermaid flowchart block representing the tree; every leaf ends in an action, never a loop. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact (including the Mermaid block), then a note on how to add a new branch without breaking the anchor links. </format>
Builds a branching decision-tree troubleshooter with anchor navigation and a Mermaid map, ready to publish.
Pro tip: List your branch questions in the order support actually asks them; Claude will mirror that flow so the tree feels like a real triage call.
Glossary & Definitions
5 promptsSingle-Term Definition Entry
11/30You are a knowledge-base editor who writes precise, standalone glossary entries. <context> I need a ready-to-publish definition article for a single product or industry term, returned as one self-contained Markdown document that could stand alone or slot into a glossary. </context> <inputs> - The term: [WORD OR PHRASE] - Domain or product it belongs to: [CONTEXT] - Plain-language meaning: [WHAT IT ACTUALLY IS] - Why it matters to the user: [WHY THEY CARE] - A concrete example of it in use: [EXAMPLE] - Terms it is often confused with: [SIMILAR TERMS] </inputs> <task> Write the entry with: an H1 of the term, a one-sentence plain-language definition first (the TL;DR), an "In more detail" paragraph, a "Why it matters" section tying it to a user outcome, a concrete real-world example in a callout, a "Not to be confused with" note distinguishing it from similar terms, and a "Related terms" list of 3 linked entries. </task> <constraints> - Valid, self-contained Markdown; lead with the plainest one-sentence definition before any nuance. - Neutral, factual tone; define any jargon used inside the definition; no marketing spin. - Include a top meta block (title, description under 160 chars, keyword = the term) as an HTML comment for SEO. </constraints> <format> Return the full entry as a Markdown artifact, then a note on how to keep the definition consistent if the term appears in other articles. </format>
Produces a standalone glossary definition entry with a TL;DR, example, and disambiguation, ready to publish.
Pro tip: Ask Claude to write the one-sentence TL;DR so a beginner gets it, then let the detail paragraph carry the nuance for advanced users.
Full A-Z Product Glossary
12/30You are a documentation lead who builds comprehensive product glossaries. <context> I need a ready-to-publish A-Z glossary of my product's key terms, returned as one self-contained Markdown document. </context> <inputs> - Product and category: [NAME + WHAT IT IS] - The terms to define (paste your list): [TERM LIST] - Audience expertise level: [BEGINNER / MIXED / EXPERT] - Terms that must link to a full article: [DEEP-DIVE TERMS] - House style for tone: [FORMAL / FRIENDLY] - Any terms to explicitly exclude or merge: [EXCLUSIONS] </inputs> <task> Write the glossary with: an H1 title, a one-line intro and a note on how it is organized, a jump-links alphabet bar (A B C ...), then alphabetized entries grouped under letter headers, each entry giving the term as a bold lead-in followed by a one-to-two-sentence definition and, where relevant, "See also" cross-references to other terms. Keep every definition consistent in length and depth. </task> <constraints> - Valid, self-contained Markdown, strictly alphabetized, with letter-group headers and an anchor jump bar. - Consistent definition style; cross-references only to terms defined in the same glossary; no duplicates. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full glossary as a Markdown artifact, then a note on how to insert a new term while keeping alphabetical order and cross-links intact. </format>
Generates a complete alphabetized product glossary with jump links and cross-references, ready to publish.
Pro tip: Paste your raw term list unsorted; Claude alphabetizes it and flags any near-duplicate terms you may want to merge.
Acronym & Abbreviation Reference
13/30You are a technical editor who documents acronyms and abbreviations for a help center. <context> I need a ready-to-publish reference article that expands and explains the acronyms used across my product, returned as one self-contained Markdown document. </context> <inputs> - Product or industry: [CONTEXT] - The acronyms to cover (paste list): [ACRONYM LIST] - Audience: [INTERNAL / CUSTOMERS / BOTH] - Any acronym with a product-specific meaning: [OVERRIDES] - Acronyms that link to a full article: [DEEP-DIVE ONES] - Tone: [PLAIN / TECHNICAL] </inputs> <task> Write the article with: an H1 title, a one-line intro, and a sorted reference table with columns for Acronym, Expansion, and What it means in one sentence, followed by short expanded notes for the 3-5 acronyms that carry a product-specific meaning different from the industry-standard one. Flag any acronym that has more than one meaning in the product. </task> <constraints> - Valid, self-contained Markdown with one alphabetized table plus expanded notes. - Every expansion is correct; product-specific meanings clearly separated from standard ones; no guessing. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to handle an acronym that means two different things in different modules. </format>
Builds an acronym and abbreviation reference table with notes on product-specific meanings, ready to publish.
Pro tip: Mark any acronym whose product meaning differs from the industry norm; Claude will call out the difference so users aren't misled.
"X vs Y" Disambiguation Article
14/30You are a knowledge-base writer who clarifies commonly confused concepts. <context> I need a ready-to-publish article that explains the difference between two similar terms or features users mix up, returned as one self-contained Markdown document. </context> <inputs> - The two things being compared: [TERM A vs TERM B] - Product context: [WHERE THEY APPEAR] - The core distinction in one sentence: [KEY DIFFERENCE] - Where each one is the right choice: [WHEN TO USE A / WHEN TO USE B] - The mistake users make by confusing them: [CONSEQUENCE] - Any overlap where both apply: [OVERLAP] </inputs> <task> Write the article with: an H1 like "[A] vs [B]: What's the difference?", a one-sentence answer up top for skimmers, a side-by-side comparison table (dimension, A, B), a "When to use [A]" and "When to use [B]" pair of sections with concrete scenarios, a "Common mix-up and why it matters" note, and a short "Quick rule of thumb" takeaway line. </task> <constraints> - Valid, self-contained Markdown with one comparison table; the one-sentence answer must appear before the table. - Neutral and precise; give concrete scenarios, not abstractions; no marketing bias toward either option. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to extend it into a three-way comparison if a related term is added. </format>
Creates a disambiguation article comparing two confused terms with a table and rule of thumb, ready to publish.
Pro tip: Give Claude the exact scenario where users pick the wrong one; it will anchor the 'when to use' sections around that real mistake.
Concept Explainer Article
15/30You are an instructional writer who explains how a system or concept works for a help center. <context> I need a ready-to-publish explainer article that teaches how one underlying concept in my product works, returned as one self-contained Markdown document. </context> <inputs> - The concept: [E.G. HOW SYNC WORKS / HOW PERMISSIONS INHERIT] - Product context: [WHERE IT APPLIES] - The mental model I want users to leave with: [KEY IDEA] - The moving parts involved: [COMPONENTS] - A relatable analogy if useful: [ANALOGY OR "NONE"] - Practical implications for the user: [WHAT IT MEANS FOR THEM] </inputs> <task> Write the article with: an H1 title, a one-paragraph plain-English overview, a "The short version" summary box, a "How it works" section that walks through the moving parts in order (with an optional analogy), a labeled Mermaid diagram of the flow or relationship, a "What this means for you" section of practical implications, and a "Common misconceptions" list correcting 2-3 wrong assumptions. </task> <constraints> - Valid, self-contained Markdown including one fenced Mermaid diagram of the concept. - Explain from the user's point of view, not the engineering internals; define terms on first use; no jargon dumps. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact (with the Mermaid diagram), then a note on how to simplify the diagram for a beginner audience. </format>
Generates a concept explainer with a mental model, Mermaid diagram, and misconceptions section, ready to publish.
Pro tip: State the one mental model you want users to walk away with; Claude will structure the whole explainer to reinforce that single idea.
Policy Articles
5 promptsRefund & Cancellation Policy
16/30You are a policy writer who turns internal rules into clear, customer-facing help-center policy articles. <context> I need a ready-to-publish refund-and-cancellation policy article for my help center, returned as one self-contained Markdown document written in plain language (not legalese). </context> <inputs> - Product and billing model: [SUBSCRIPTION / ONE-TIME / USAGE] - Refund eligibility window and conditions: [E.G. 14 DAYS, UNUSED] - What is non-refundable: [EXCLUSIONS] - How to cancel and when it takes effect: [CANCEL PROCESS] - What happens to data and access after cancellation: [DATA POLICY] - How refunds are issued and timing: [METHOD + TIMEFRAME] </inputs> <task> Write the policy with: an H1 title, an effective-date line, a plain-language summary of the policy in 2-3 sentences, a "Am I eligible for a refund?" section with clear yes/no conditions, a "What isn't refundable" list, a step-by-step "How to cancel" section, a "What happens after you cancel" section covering access and data, a "How and when you'll get your refund" section, and a "Contact us" footer. Keep it accurate and unambiguous. </task> <constraints> - Valid, self-contained Markdown in plain language; every rule stated as a concrete condition, no vague "may". - Include an effective date and a note that terms may update; do not invent legal guarantees not in the inputs. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full policy as a Markdown artifact, then a note flagging any place where you had to assume a rule the inputs didn't specify. </format>
Produces a plain-language refund and cancellation policy article with clear eligibility rules, ready to publish.
Pro tip: Ask Claude to flag every rule it had to assume; those flags are exactly what your legal or ops team should confirm before publishing.
Privacy & Data-Handling Explainer
17/30You are a privacy-focused writer who explains data practices to customers in plain terms. <context> I need a ready-to-publish help-center article explaining how my product handles user data, returned as one self-contained Markdown document. This is a customer-friendly explainer that complements (not replaces) the legal privacy policy. </context> <inputs> - Product: [NAME] - What data you collect and why: [DATA TYPES + PURPOSE] - Where it is stored and for how long: [STORAGE + RETENTION] - Who it is shared with, if anyone: [SUBPROCESSORS / NONE] - User controls available: [EXPORT, DELETE, OPT-OUT] - Relevant frameworks: [GDPR / CCPA / SOC 2 / NONE] </inputs> <task> Write the article with: an H1 title, a one-paragraph plain-language summary, a "What we collect and why" table (data type, purpose, retention), a "Where your data lives" section, a "Who can access it" section covering any third parties, a "Your controls" section explaining how to export or delete data with steps, a compliance note referencing applicable frameworks, and a link-out line to the full legal privacy policy. </task> <constraints> - Valid, self-contained Markdown with a data table; plain language; clearly labeled as an explainer, not the legal policy. - Do not overstate compliance beyond the inputs; give real steps for user data controls; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on which claims must be verified against your actual legal privacy policy before publishing. </format>
Generates a customer-friendly privacy and data-handling explainer with a data table and user controls, ready to publish.
Pro tip: Keep this as a plain-language companion to your legal policy; tell Claude to link out rather than restate binding legal terms.
Acceptable Use Policy Article
18/30You are a trust-and-safety writer who documents acceptable use rules for a product. <context> I need a ready-to-publish acceptable-use policy article for my help center, returned as one self-contained Markdown document that users can actually understand. </context> <inputs> - Product and what people do with it: [USE CONTEXT] - Behaviors that are prohibited: [PROHIBITED LIST] - Content or actions that are restricted or limited: [RESTRICTIONS] - Consequences of violations: [WARNING / SUSPENSION / BAN] - How to report a violation: [REPORTING CHANNEL] - The appeal process, if any: [APPEALS] </inputs> <task> Write the policy with: an H1 title, an effective-date line, a short plain-language intro on the spirit of the policy, a "What's not allowed" section with grouped, concrete examples for each prohibited category, a "What happens if you break these rules" enforcement-ladder section, a "How to report abuse" section, an "Appealing a decision" section, and a contact footer. Use concrete examples so rules aren't ambiguous. </task> <constraints> - Valid, self-contained Markdown; each prohibited category has at least one concrete example of what does and doesn't cross the line. - Fair, non-threatening tone; enforcement steps stated as a clear ladder; do not invent consequences not in the inputs. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full policy as a Markdown artifact, then a note on where a concrete example would most reduce edge-case disputes. </format>
Builds an acceptable-use policy article with concrete examples and an enforcement ladder, ready to publish.
Pro tip: For each prohibited category, ask Claude for a 'crosses the line vs doesn't' example pair so moderators and users read the rule the same way.
SLA & Support Policy Article
19/30You are a technical writer documenting service-level and support commitments. <context> I need a ready-to-publish article explaining my product's uptime commitment and support response policy, returned as one self-contained Markdown document. </context> <inputs> - Product and plans covered: [PLANS] - Uptime target and how it is measured: [E.G. 99.9%, MONTHLY] - Support channels and hours by plan: [CHANNELS + HOURS] - Response-time targets by priority level: [P1..P4 TARGETS] - What counts as excluded downtime: [MAINTENANCE, FORCE MAJEURE] - Remedies or credits for missed targets: [SLA CREDITS] </inputs> <task> Write the article with: an H1 title, an effective-date line, a plain-language summary, an "Uptime commitment" section defining the target and how it's measured, a support-response table (priority level, definition, target response, channel), a "Support hours by plan" table, an "Exclusions" section for scheduled maintenance and events outside the target, a "Service credits" section explaining how to claim, and a status-page link line. </task> <constraints> - Valid, self-contained Markdown with two tables (response targets and support hours by plan). - Define each priority level with a concrete example; do not promise targets beyond the inputs; state measurement window explicitly. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to keep the plan tables in sync when a new plan is added. </format>
Creates an SLA and support-policy article with priority and plan tables plus credit terms, ready to publish.
Pro tip: Define each priority level with a real example incident; vague P1/P2 labels are the top source of SLA disputes.
Fair-Use & Rate-Limit Policy
20/30You are a developer-documentation writer who documents usage limits and fair-use rules. <context> I need a ready-to-publish policy article explaining my product's usage limits, rate limits, and fair-use rules, returned as one self-contained Markdown document. </context> <inputs> - Product or API: [NAME] - The limits by plan: [REQUESTS/MIN, STORAGE, SEATS, ETC] - How limits are counted and reset: [WINDOW + RESET LOGIC] - What happens when a limit is hit: [THROTTLE / 429 / BLOCK] - How to request higher limits: [PROCESS] - The fair-use principle behind soft limits: [RATIONALE] </inputs> <task> Write the article with: an H1 title, a plain-language intro on why limits exist, a limits-by-plan table (metric, and value per plan), a "How limits are counted" section explaining the window and reset, a "What happens when you hit a limit" section (response code, headers, retry guidance) with a sample response in a code block, a "Fair use" section explaining soft limits, and a "How to request an increase" section. </task> <constraints> - Valid, self-contained Markdown with a per-plan limits table and a sample rate-limit response in a fenced code block. - Be specific about counting windows and reset behavior; give concrete retry guidance; no vague "reasonable use" without definition. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to document a new limit tier without confusing existing customers. </format>
Generates a fair-use and rate-limit policy with a per-plan table and sample 429 response, ready to publish.
Pro tip: Include the real rate-limit response headers so developers can build retry logic straight from the article without opening a ticket.
FAQ-Style Articles
5 promptsProduct FAQ Page
21/30You are a support-content writer who builds skimmable FAQ pages that deflect tickets. <context> I need a ready-to-publish FAQ article covering the most common questions about my product, returned as one self-contained Markdown document. </context> <inputs> - Product and what it does: [ONE-LINER] - The real questions users ask (paste them): [QUESTION LIST] - Logical groups for the questions: [GROUPS, E.G. GETTING STARTED, BILLING] - Answers or key facts for each: [SOURCE FACTS] - Questions that should link to a full article: [DEEP-DIVE ONES] - Tone: [FRIENDLY / PROFESSIONAL] </inputs> <task> Write the FAQ with: an H1 title, a one-line intro, grouped H2 sections, and under each a series of question-as-H3 entries followed by a tight 2-4 sentence answer. Lead each answer with the direct response, then any caveat. For questions that warrant depth, end the answer with a "Learn more" link to the relevant article. Order questions within each group from most to least common. </task> <constraints> - Valid, self-contained Markdown; each question phrased the way a real user would ask it; answers lead with the direct answer. - No question left unanswered or answered with only a link; concise, no filler; consistent answer length. - Include a top meta block (title, description under 160 chars) plus FAQPage JSON-LD in a fenced code block for rich results. </constraints> <format> Return the full FAQ as a Markdown artifact (including the JSON-LD block), then a note on how to prune or reorder questions as ticket trends change. </format>
Produces a grouped product FAQ with direct answers and FAQ schema markup, ready to publish.
Pro tip: Paste the exact wording of questions from your inbox; matching real phrasing is what makes the FAQ rank and actually get found.
Billing & Pricing FAQ
22/30You are a support writer who specializes in clear billing and pricing answers. <context> I need a ready-to-publish billing-and-pricing FAQ for my help center, returned as one self-contained Markdown document. </context> <inputs> - Billing model and plans: [SUBSCRIPTION / USAGE, PLAN NAMES] - Billing cycles and proration rules: [MONTHLY/ANNUAL, PRORATION] - Payment methods and invoicing: [CARDS, ACH, INVOICE] - Common billing questions users have: [QUESTION LIST] - Taxes, currency, and receipts: [TAX/VAT, CURRENCY] - Upgrade, downgrade, and refund basics: [PLAN-CHANGE RULES] </inputs> <task> Write the FAQ with: an H1 title, a one-line intro, and grouped sections for Plans & pricing, Payments & invoices, Changing your plan, and Taxes & receipts. Under each, write real questions as H3s with direct 2-4 sentence answers, including how proration works, when charges hit, how to update payment details, and how to get an invoice or VAT receipt. Cross-link the refund policy where relevant. </task> <constraints> - Valid, self-contained Markdown; money and timing facts must be exact and unambiguous. - Explain proration and cycle timing with a concrete example; do not invent fees not in the inputs. - Include a top meta block (title, description under 160 chars) plus FAQPage JSON-LD in a fenced code block. </constraints> <format> Return the full FAQ as a Markdown artifact (with JSON-LD), then a note on which billing question most often becomes a ticket and how the answer preempts it. </format>
Generates a billing and pricing FAQ with proration examples and schema markup, ready to publish.
Pro tip: Give Claude a real proration scenario with numbers; a worked example deflects far more tickets than a policy sentence.
New-User Onboarding FAQ
23/30You are a customer-education writer who answers the questions new users ask in week one. <context> I need a ready-to-publish FAQ aimed at brand-new users, returned as one self-contained Markdown document that reduces early confusion and churn. </context> <inputs> - Product and first-use goal: [WHAT NEW USERS WANT TO DO] - The questions new users ask first: [EARLY QUESTIONS] - Setup blockers they hit: [COMMON BLOCKERS] - Free-trial or plan details relevant early on: [TRIAL RULES] - Where to get hands-on help: [SUPPORT / ONBOARDING] - Quick wins to point them to: [FAST VALUE ACTIONS] </inputs> <task> Write the FAQ with: an H1 title, a warm one-line intro, and grouped sections for Getting set up, Doing your first task, Trial & plans, and Getting help. Under each, phrase real beginner questions as H3s with reassuring, direct answers, linking each answer to the relevant getting-started article or quick win. Prioritize the questions that block a user from reaching first value. </task> <constraints> - Valid, self-contained Markdown; answers reassure without condescending; each links to a next step where useful. - Front-load the questions that block activation; concise answers; no jargon without a one-line gloss. - Include a top meta block (title, description under 160 chars) plus FAQPage JSON-LD in a fenced code block. </constraints> <format> Return the full FAQ as a Markdown artifact (with JSON-LD), then a note on how to keep it aligned with your onboarding flow when the product changes. </format>
Creates a new-user onboarding FAQ focused on early questions and activation blockers, ready to publish.
Pro tip: Ask Claude to order questions by what blocks activation first; answering the blocker before the nice-to-know keeps new users moving.
Security & Compliance FAQ
24/30You are a writer who answers security and compliance questions for prospects and customers. <context> I need a ready-to-publish security-and-compliance FAQ for my help center, returned as one self-contained Markdown document that answers the questions buyers and IT teams ask. </context> <inputs> - Product: [NAME] - Certifications and audits held: [SOC 2, ISO 27001, ETC OR NONE] - Data encryption at rest and in transit: [DETAILS] - Access controls and authentication: [SSO, 2FA, RBAC] - Data residency and backup practices: [REGIONS, BACKUPS] - Where to request security docs: [DPA, REPORTS, CONTACT] </inputs> <task> Write the FAQ with: an H1 title, a one-line intro, and grouped sections for Certifications, Data protection, Access & authentication, and Data residency & continuity. Phrase real IT-buyer questions as H3s with precise, verifiable answers. Include how encryption works, what auth options exist, and how to request a DPA or audit report. Keep answers factual and avoid overclaiming. </task> <constraints> - Valid, self-contained Markdown; every claim must be truthful and match the inputs, with no compliance overstatement. - Precise, confident, non-marketing tone; point to where documents can be requested rather than pasting confidential details. - Include a top meta block (title, description under 160 chars) plus FAQPage JSON-LD in a fenced code block. </constraints> <format> Return the full FAQ as a Markdown artifact (with JSON-LD), then a note listing any answer that must be verified by your security team before publishing. </format>
Builds a security and compliance FAQ for IT buyers with verifiable, non-overstated answers, ready to publish.
Pro tip: Have Claude flag every security claim it wrote; those flags are the exact list your security team should sign off before you go live.
Pre-Sales "Before You Buy" FAQ
25/30You are a conversion-minded support writer who answers the objections prospects have before purchasing. <context> I need a ready-to-publish pre-sales FAQ that answers the questions people ask before buying, returned as one self-contained Markdown document. </context> <inputs> - Product and who it's for: [PRODUCT + AUDIENCE] - The buying objections and questions prospects raise: [OBJECTION LIST] - What makes you different from alternatives: [DIFFERENTIATORS] - Trial, guarantee, and cancellation basics: [RISK REVERSAL] - Migration or onboarding support offered: [SWITCHING HELP] - Where to talk to sales or start: [CTA LINKS] </inputs> <task> Write the FAQ with: an H1 title, a one-line intro, and grouped sections for Is it right for me, How it compares, Getting started risk-free, and Switching & support. Phrase real prospect questions as H3s with honest, confidence-building answers that address the objection directly, then reduce risk (trial, guarantee) and point to a clear next step. Be persuasive through clarity, not hype. </task> <constraints> - Valid, self-contained Markdown; answer the actual objection first, then reassure; no dodging hard questions. - Honest about fit (including who it's NOT for); avoid slop words like "revolutionary"; concrete differentiators only. - Include a top meta block (title, description under 160 chars) plus FAQPage JSON-LD in a fenced code block. </constraints> <format> Return the full FAQ as a Markdown artifact (with JSON-LD), then a note on which objection, if answered better, would most improve conversion. </format>
Generates a pre-sales FAQ that answers buying objections honestly with risk reversal, ready to publish.
Pro tip: Include a straight answer to 'who is this NOT for'; disqualifying honestly builds more trust than pretending you fit everyone.
Most people use 10% of Claude. Tutorials unlock the rest.
AI Academy: 300+ hands-on tutorials on Claude, ChatGPT, Midjourney, and 50+ AI tools. New tutorials added every week.
Feature Guides
5 promptsSingle-Feature Deep-Dive Guide
26/30You are a product-education writer who creates thorough feature guides for a help center. <context> I need a ready-to-publish deep-dive guide for one product feature, returned as one self-contained Markdown document that takes a user from "what is it" to confident use. </context> <inputs> - Feature name and what it does: [FEATURE + PURPOSE] - The problem it solves for the user: [JOB TO BE DONE] - How to access and enable it: [LOCATION / TOGGLE] - Its main settings or options: [KEY OPTIONS] - Plan or permission requirements: [ACCESS RULES] - Best-practice tips and limitations: [TIPS + LIMITS] </inputs> <task> Write the guide with: an H1 feature name, a one-line "what it does" summary, a "When to use it" section with real scenarios, a "How to turn it on" numbered section, a "Key settings explained" section covering each option and what it changes, a short worked example walkthrough, a "Tips & best practices" list, a "Limitations & things to know" list, and a "Related features" links list. </task> <constraints> - Valid, self-contained Markdown; explain each setting in terms of the outcome it produces, not just its label. - Include one concrete worked example; exact UI labels in bold; state plan requirements up front; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full guide as a Markdown artifact, then a note on where a short demo GIF or video would add the most value. </format>
Produces a single-feature deep-dive with settings explained and a worked example, ready to publish.
Pro tip: For each setting, tell Claude the outcome it changes; users need to know what a toggle does for them, not just what it's called.
Feature-by-Plan Comparison Guide
27/30You are a documentation writer who clarifies which features come with which plan. <context> I need a ready-to-publish article that explains what features are included in each plan and how they differ, returned as one self-contained Markdown document. </context> <inputs> - Plan names in order: [E.G. FREE, PRO, BUSINESS, ENTERPRISE] - The features to compare and per-plan availability: [FEATURE MATRIX DATA] - Usage limits that differ by plan: [SEATS, STORAGE, LIMITS] - Features that are add-ons vs included: [ADD-ONS] - Who each plan is best suited for: [PLAN PERSONAS] - How to upgrade or compare live pricing: [LINK] </inputs> <task> Write the article with: an H1 title, a one-line intro, a "Which plan is right for you" quick-guidance block mapping each plan to a persona, a full feature-comparison table (feature rows, plan columns, using check/dash/limit values), a short section per plan summarizing what you get and who it's for, a "Feature limits explained" section for any metered items, and an "Add-ons" section for anything sold separately. </task> <constraints> - Valid, self-contained Markdown with one clean comparison table; use consistent symbols (included / not included / limited) with a legend. - Match every value to the inputs; don't imply a feature exists on a plan where it doesn't; note where limits apply. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to update the table when a feature moves between plans. </format>
Builds a feature-by-plan comparison guide with a matrix table and persona guidance, ready to publish.
Pro tip: Give Claude a legend for your symbols (check, dash, or a number); a consistent legend keeps a big feature matrix readable.
"What's New" Feature Release Explainer
28/30You are a product writer who turns release notes into a customer-facing feature explainer. <context> I need a ready-to-publish "what's new" article that explains a recently shipped feature and helps users adopt it, returned as one self-contained Markdown document. </context> <inputs> - Feature name and release date: [NAME + DATE] - What it does and the problem it solves: [VALUE] - What changed for existing users: [BEHAVIOR CHANGES] - How to start using it: [ENABLE / ACCESS STEPS] - Who it's available to: [PLANS / ROLLOUT STATUS] - Where to give feedback: [FEEDBACK CHANNEL] </inputs> <task> Write the article with: an H1 announcing the feature, a release-date line, a punchy "What's new" summary paragraph, a "Why we built it" short block tied to a user problem, a "How to use it" numbered section, a "What this changes for you" note for existing users (including any migration or default changes), an availability note (plans and rollout), and a "Tell us what you think" feedback footer. </task> <constraints> - Valid, self-contained Markdown; benefit-led but factual; clearly separate new behavior from what stayed the same. - Call out any change to defaults or existing workflows explicitly so nobody is surprised; avoid hype words. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full article as a Markdown artifact, then a note on how to trim it into a shorter in-app changelog entry. </format>
Generates a customer-facing feature-release explainer with adoption steps and change notes, ready to publish.
Pro tip: Tell Claude exactly what changed for existing users; a clear 'what this changes for you' block prevents the surprise tickets after a release.
Feature Use-Case / Workflow Guide
29/30You are a solutions writer who documents end-to-end workflows using product features. <context> I need a ready-to-publish guide showing how to accomplish a real workflow using one or more features together, returned as one self-contained Markdown document. </context> <inputs> - The workflow or outcome: [E.G. RUN A WEEKLY REPORT] - The features involved and their order: [FEATURE SEQUENCE] - The user role and their goal: [WHO + WHY] - Inputs or data needed to start: [PREREQUISITES] - The end result they'll have: [DELIVERABLE] - Tips to make it repeatable: [AUTOMATION / TEMPLATES] </inputs> <task> Write the guide with: an H1 outcome-phrased title, a "What you'll accomplish" summary, a "Before you start" prerequisites list, an end-to-end numbered walkthrough that chains the features in order (bolding the exact actions and naming which feature is being used at each step), a "What you should see" result section, a "Make it repeatable" section (templates, saved views, or automation), and a "Troubleshooting the workflow" mini-section for the 2-3 spots people get stuck. </task> <constraints> - Valid, self-contained Markdown; steps must chain logically so the reader ends with the stated deliverable. - Name the feature used at each step so users learn the tool, not just the clicks; exact UI labels in bold; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full guide as a Markdown artifact, then a note on how to adapt the workflow for a different role or a larger dataset. </format>
Creates an end-to-end workflow guide that chains features toward a concrete deliverable, ready to publish.
Pro tip: Describe the exact deliverable a user should end with; Claude reverse-engineers the steps so the workflow actually produces it.
Settings & Configuration Reference
30/30You are a technical writer who documents settings and configuration options for a help center. <context> I need a ready-to-publish reference article documenting a settings area of my product, returned as one self-contained Markdown document users can scan to find any option. </context> <inputs> - The settings area covered: [E.G. NOTIFICATION SETTINGS] - The individual settings, their options, and defaults: [SETTING LIST] - What each setting actually changes: [EFFECT PER SETTING] - Which settings need admin rights: [ADMIN-ONLY ONES] - Recommended defaults for common cases: [RECOMMENDATIONS] - Settings people commonly misconfigure: [GOTCHAS] </inputs> <task> Write the reference with: an H1 title, a one-line intro and how to reach these settings, a settings reference table (setting name, options, default, what it controls), then expanded H3 entries for the settings that need more explanation or carry a warning, a "Recommended setup" section for the most common use case, and a "Common misconfigurations" section explaining the 2-3 settings people get wrong and the symptom each causes. </task> <constraints> - Valid, self-contained Markdown with a settings table including defaults; flag admin-only settings clearly. - Describe each setting by its real-world effect, not a restatement of its label; no filler. - Include a top meta block (title, description under 160 chars) as an HTML comment. </constraints> <format> Return the full reference as a Markdown artifact, then a note on how to add a new setting row without disrupting the table or the recommended-setup section. </format>
Builds a settings and configuration reference with a defaults table and misconfiguration warnings, ready to publish.
Pro tip: Tell Claude which settings people most often get wrong; the 'common misconfigurations' section quietly deflects a whole class of tickets.
Frequently Asked Questions
Prompts are the starting line. Tutorials are the finish.
A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.
7-day free trial. Cancel anytime.