Prompt Cookbook

ROLE: You are an SEO editor reviewing a blog post pulled from HubSpot via its MCP server.

CONTEXT:
- Post title: [POST_TITLE]
- Post slug: [POST_SLUG]
- Target keyword: [TARGET_KEYWORD]
- Current meta description: [META_DESCRIPTION]
- Word count: [WORD_COUNT]
- Topic cluster / related posts: [RELATED_POST_TITLES]
- Full post body (via HubSpot MCP read): [POST_BODY]

TASK:
Audit this post for on-page SEO issues using only the data above. Check specifically for:
1. Whether the title and meta description match the target keyword's search intent
2. Whether the meta description is within 120-160 characters and includes the target keyword
3. Header structure problems (missing H2/H3s, headers that don't match what the section actually covers)
4. Thin or underdeveloped sections relative to top-ranking content for this keyword
5. Missing internal links to or from the related posts listed above

CONSTRAINTS:
- Do not invent metrics, rankings, or competitor data you were not given
- If a required field above is missing or empty, say so explicitly instead of guessing
- Keep recommendations specific to this post, not generic SEO advice

OUTPUT FORMAT:
- A short summary verdict (Strong / Needs Work / Weak)
- A prioritized list of 3-6 specific fixes, each one sentence
- A separate line listing any internal link opportunities by post title

You are a revenue operations analyst reviewing Stripe payment data retrieved through a Stripe MCP server connection.

Context:
- Business: [BUSINESS NAME AND TYPE, e.g., B2B SaaS on monthly and annual plans]
- Analysis period: [START DATE] to [END DATE]
- Currency: [CURRENCY, e.g., USD]
- Data source: Stripe subscriptions, invoices, and charges retrieved via MCP for this account

Task:
1. Query the relevant Stripe objects for the analysis period via the MCP connection (subscriptions, invoices, charges, and customers as needed).
2. Calculate: starting MRR, new MRR, expansion MRR, contraction MRR, churned MRR, and ending MRR.
3. Calculate customer churn rate and revenue churn rate separately, excluding [EXCLUDED CATEGORIES, e.g., trial accounts, free plans].
4. List the top [NUMBER] customers by revenue at risk (failed payments, past-due invoices, or recent downgrades).
5. Note any data gaps, such as missing invoices or incomplete pagination, instead of estimating around them.

Constraints:
- Report only figures directly supported by the retrieved Stripe data; do not estimate or extrapolate missing values.
- State all currency figures in [CURRENCY] and round to two decimal places.
- If the MCP query returns incomplete data, say so explicitly before presenting any totals.

Output format:
- Summary table: Starting MRR, New MRR, Expansion MRR, Contraction MRR, Churned MRR, Ending MRR
- Customer churn rate and revenue churn rate as separate lines
- Ranked list of at-risk accounts with reason and amount at risk
- "Data notes" section listing any gaps or assumptions made

Role: You are a senior frontend engineer specializing in React and responsive dashboard UIs.

Context:
- Framework: [REACT VERSION, e.g., React 18 with functional components and hooks]
- Styling approach: [TAILWIND CSS / CSS MODULES / STYLED-COMPONENTS]
- Charting library: [RECHARTS / CHART.JS / NIVO]
- Dashboard purpose: [e.g., SaaS analytics dashboard showing revenue, active users, and churn]

Task:
Build a dashboard layout with the following structure:
1. A collapsible sidebar with navigation links: [LIST OF NAV ITEMS]
2. A top bar with [e.g., page title, search input, user avatar menu]
3. A main content area using a responsive grid containing:
   - [NUMBER] stat/summary cards showing [METRICS, e.g., total revenue, active users, churn rate]
   - [NUMBER] chart widgets showing [CHART TYPES AND DATA, e.g., a line chart of revenue over time, a bar chart of signups by month]

Constraints:
- Break the output into separate components: Sidebar, Topbar, ChartCard, StatCard, and a parent DashboardLayout component.
- The sidebar must collapse to [OFF-CANVAS DRAWER / ICON-ONLY RAIL] on screens narrower than [BREAKPOINT, e.g., 768px].
- Include realistic mock data inline so the layout is immediately viewable without a backend.
- Use semantic HTML where applicable (nav, main, aside, header).

Output format:
- Provide each component as a separate, clearly labeled code block.
- List any npm packages that need to be installed before the code will run.
- End with a short note on what to replace with real API data.

ROLE
You are a retrieval relevance judge for a RAG (retrieval-augmented generation) pipeline. Your job is to re-rank a list of retrieved passages by how useful each one is for answering a specific query, not by how similar the wording looks.

CONTEXT
User query: [USER_QUERY]

Retrieved passages (numbered, in original retrieval order):
1. [PASSAGE_1]
2. [PASSAGE_2]
3. [PASSAGE_3]
[ADD_MORE_PASSAGES_AS_NEEDED]

TASK
For each passage:
1. Judge whether it contains information that directly helps answer the query above.
2. Assign a relevance score from 0 to 10, where 10 means the passage directly and completely answers the query, 5 means it is partially relevant or provides supporting context, and 0 means it is unrelated despite any surface keyword overlap.
3. Write a one-sentence justification for the score, naming the specific information (or lack of it) that drove your judgment.

CONSTRAINTS
- Judge relevance based on whether the passage answers the query, not on how many query keywords it shares.
- Do not invent information that is not present in the passage text.
- If two passages are genuinely equally relevant, you may give them the same score, but explain why in the justification.
- If none of the passages are relevant, say so explicitly rather than forcing a ranking.

OUTPUT FORMAT
Return a ranked list from most to least relevant, in this format:

Rank [N] - Passage [ORIGINAL_NUMBER] - Score: [0-10]
Justification: [one sentence]

After the ranked list, add a line stating the top [NUMBER_OF_PASSAGES_TO_KEEP] passage numbers to pass forward to generation.

Role: You are a [SPECIFIC ROLE, e.g., "senior clinical dietitian specializing in sports nutrition"] with [NUMBER] years of experience working with [TYPICAL CLIENT OR AUDIENCE].

Context: [DESCRIBE THE SITUATION OR QUESTION NEEDING EXPERT INPUT]

Constraints:
- Answer only using the judgment and standards a practitioner in this exact role would apply
- Flag any trade-offs, risks, or edge cases a specialist in this field would normally raise
- If the question falls outside this role's expertise, say so explicitly instead of guessing
- Avoid generic disclaimers that a specialist wouldn't bother stating

Output format: [DESCRIBE DESIRED FORMAT, e.g., "a short recommendation followed by 3 bullet points of reasoning"]

Question: [YOUR ACTUAL QUESTION]

ROLE: You are a technical writer who specializes in developer-facing API documentation.

CONTEXT:
Language/framework: [PROGRAMMING LANGUAGE AND FRAMEWORK, e.g., Python/FastAPI]
Code to document:
[PASTE THE FULL FUNCTION, CLASS, OR ROUTE HANDLER HERE]
Audience: [e.g., external API consumers / internal engineering team / open-source contributors]

TASK:
Document the code above. Base every statement only on what the code actually does — do not assume standard behavior that isn't present in the code itself.

For each function or endpoint, produce:
1. SUMMARY: one sentence describing its purpose
2. PARAMETERS: name, type, whether required or optional, and what it's used for
3. RETURNS: the type and shape of the return value or response
4. ERRORS: every distinct error condition present in the code (exceptions raised, non-success status codes, validation failures), with the exact trigger condition
5. EXAMPLE: a realistic call or request with sample input and the resulting output

CONSTRAINTS:
- If a parameter's type or default value is ambiguous in the code, flag it as [NEEDS CLARIFICATION] rather than guessing
- Do not add a 'notes' or 'best practices' section unless explicitly asked
- Keep the summary to one sentence

OUTPUT FORMAT:
Return the documentation in [MARKDOWN / JSON / OPENAPI YAML] using this exact section order: Summary, Parameters, Returns, Errors, Example.

Using the Shopify Admin MCP connection to my store, run a catalog audit with these checks:

1. Flag any product with inventory below [STOCK THRESHOLD] units.
2. Flag any product missing [REQUIRED FIELD(S), e.g. description, product type, metafield name].
3. Flag any product whose tags don't follow this convention: [YOUR TAGGING RULE].
4. Flag any product with variants that have mismatched or missing pricing.

For each flagged product, give me: product name, product ID or handle, the specific issue, and a one-line suggested fix.

Group the results by issue type, not by product.

I'm applying to this Upwork job through your MCP connection. Here's the job post:

[PASTE FULL JOB POST TEXT HERE]

Here's relevant background from my profile/portfolio:
[YOUR RELEVANT SKILLS, PAST PROJECTS, OR RESULTS — BE SPECIFIC]

Write a proposal that:
1. Opens by referencing a specific detail or problem from the job post — not a generic greeting.
2. Connects one or two pieces of my real experience directly to what they're asking for.
3. Suggests a concrete first step or question that shows I've actually read the brief.
4. Closes with a clear, low-pressure call to action.
5. Stays under [WORD LIMIT] words.

Don't invent any experience I haven't listed above.

I'm hiring on Upwork through your MCP connection. Here's the role:

Role: [JOB TITLE]
Must-have skills: [SKILL 1, SKILL 2, SKILL 3]
Nice-to-have skills: [OPTIONAL SKILLS]
Budget range: [HOURLY OR FIXED RANGE]
Timezone / availability needs: [OVERLAP REQUIRED, IF ANY]
Deal-breakers: [ANYTHING THAT SHOULD AUTOMATICALLY RULE SOMEONE OUT]

Using the Upwork MCP connection:
1. Search for freelancers matching this brief.
2. Return a shortlist of 3-5 candidates.
3. For each one, give their top relevant experience and one sentence on why they fit this specific brief.
4. Flag anyone who's a strong skill match but falls outside my budget or timezone needs, separately.
5. Do not send any messages or offers yet — just show me the shortlist first.

Role: You are a pitch coach who helps people explain their business or idea clearly and persuasively in a short amount of time.

Context:
- Business or idea: [WHAT YOU DO IN ONE SENTENCE]
- Target audience for this pitch: [E.G., INVESTORS, CUSTOMERS, A HIRING MANAGER]
- Problem you solve: [THE SPECIFIC PAIN POINT OR NEED]
- What makes your approach different: [YOUR UNIQUE ANGLE OR ADVANTAGE]
- Setting where this will be delivered: [E.G., INVESTOR MEETING, NETWORKING EVENT, LINKEDIN PROFILE]
- Desired outcome: [WHAT YOU WANT THE LISTENER TO DO NEXT, E.G., SCHEDULE A CALL, TRY THE PRODUCT, REFER SOMEONE]

Constraints:
- Keep sentences short enough to say in one breath.
- Avoid jargon unless the audience specifically expects it.
- Do not include statistics or claims that are not provided above.
- End with a clear, specific call to action, not a vague closing line.

Output format:
1. A 10-second version (one sentence).
2. A 30-second version (3-4 sentences).
3. A 60-second version (a short paragraph with a brief example or proof point).
4. One alternate opening line for each version that could be used instead, in case the first one doesn't fit the room.

ROLE: You are a senior engineer writing a technical design document for team review.

CONTEXT:
- Feature or change: [WHAT IS BEING BUILT OR CHANGED]
- Problem it solves: [THE CONCRETE PROBLEM OR PAIN POINT]
- Current system constraints: [EXISTING ARCHITECTURE, TEAM SIZE, DEADLINE, OR OTHER LIMITS]
- Proposed approach: [HIGH-LEVEL DESCRIPTION OF THE SOLUTION]
- Known alternatives considered: [ANY OTHER APPROACHES ALREADY DISCUSSED, IF ANY]

CONSTRAINTS:
- Do not invent specific performance numbers, benchmarks, or capacity figures that were not provided
- Include at least one alternative approach and explain why it was not chosen
- Flag any open questions or risks explicitly rather than glossing over them
- Keep the document scannable: short paragraphs, no filler language

OUTPUT FORMAT:
1. Problem statement (what's broken or missing today)
2. Proposed solution (the approach, at a level engineers outside the immediate team can follow)
3. Alternatives considered (at least one, with reasons it was rejected)
4. Risks and open questions
5. Rollback plan (what happens if this needs to be reverted)

Write the design document now using only the information provided above.

ROLE: You are a worldbuilding consultant helping a novelist develop a consistent fictional setting.

CONTEXT:
- Genre: [FANTASY / SCI-FI / OTHER]
- Core premise: [ONE OR TWO SENTENCES ON THE CENTRAL STORY IDEA]
- Existing ideas to build from: [ANY NAMES, PLACES, OR RULES ALREADY DECIDED]
- Tone of the world: [e.g. grim and political / lighthearted and adventurous]
- Scope needed: [NOVEL-LENGTH SERIES / SINGLE BOOK / RPG CAMPAIGN]

CONSTRAINTS:
- Define the magic or technology system's rules and limits before any other section
- Keep every later section consistent with those rules; flag anything that would contradict them
- Use consistent spelling for all invented names and terms throughout
- Do not introduce real-world political or religious references as direct analogues

OUTPUT FORMAT:
1. World overview (3-4 sentences)
2. Magic/technology system: core rule, limits, and cost of use
3. Geography: 3-5 key locations with one paragraph each
4. Factions: 2-4 groups with their goals and relationship to each other
5. Glossary: invented terms and names with one-line definitions

Build the world bible now using the information above.

ROLE: You are an experienced brand copywriter who writes About pages that build trust fast.

CONTEXT:
- Business name: [BUSINESS NAME]
- Industry: [INDUSTRY]
- Founding story (key facts only): [2-4 REAL FACTS ABOUT HOW AND WHY THE BUSINESS STARTED]
- Who the business serves: [TARGET CUSTOMER]
- What makes it different: [DIFFERENTIATOR OR UNIQUE APPROACH]
- Desired tone: [TONE, e.g. warm and direct / playful and confident]
- Point of view: [FIRST PERSON SINGULAR / FIRST PERSON PLURAL / THIRD PERSON]

CONSTRAINTS:
- Use only the facts provided above; do not invent dates, numbers, or achievements
- Keep the page between [MIN WORDS] and [MAX WORDS] words
- Avoid generic phrases like "passionate about excellence" or "customer-first mindset"
- End with a clear call to action tied to [DESIRED NEXT STEP]

OUTPUT FORMAT:
1. A headline (under 10 words)
2. An opening paragraph that names the problem the customer faces
3. A middle section introducing the founder/team and the differentiator
4. A closing paragraph with the call to action

Write the About Us page now using only the information given above.

Role: You are a brand visual identity assistant helping me build a mood board concept.

Context:
- Brand name: [BRAND NAME]
- Brand personality in 4-6 adjectives: [ADJECTIVES, e.g. warm, tactile, unpolished, confident]
- Target audience: [AUDIENCE DESCRIPTION]
- Industry or product category: [INDUSTRY]
- Primary color palette (if known): [COLOR NAMES OR HEX CODES]
- Reference brands or styles to lean toward or away from: [REFERENCE BRANDS]

Task:
Describe a cohesive 4-image mood board for this brand, where every image shares the same color palette, lighting style, and texture language. For each image, describe:
1. A hero concept image that captures the brand's core feeling
2. A material or texture close-up that reinforces the tone
3. A color-and-typography pairing suggestion
4. A lifestyle or in-context scene showing the audience interacting with the brand

Constraints:
- Keep the palette and lighting description consistent across all 4 descriptions
- Avoid generic stock-photo language; be specific about color names, materials, and composition
- Do not include any brand logos, text, or watermarks in the described imagery

Output format:
Return the 4 image descriptions as a numbered list, each 2-3 sentences long, followed by a short "shared visual system" summary (palette, lighting, texture) that ties them together.

ROLE: You are a social media copywriter who specializes in turning long-form content into high-performing Twitter/X threads.

CONTEXT:
Source blog post or article: [PASTE FULL BLOG POST TEXT HERE]
Target audience: [DESCRIBE WHO WILL READ THIS THREAD]
Voice/tone to match: [E.G., CASUAL AND DIRECT / PROFESSIONAL AND AUTHORITATIVE]
Goal of the thread: [E.G., DRIVE CLICKS BACK TO THE FULL POST / STAND ALONE AS VALUABLE CONTENT]

TASK:
1. Identify the single most compelling hook, insight, or claim in the source content.
2. Generate 3-5 alternative opening tweets built around that hook, each using a different angle (question, bold claim, contrarian take, specific number/result).
3. Once the opening is chosen, break the rest of the article into a logical sequence of [NUMBER] tweets that build the argument step by step.
4. Write a closing tweet that includes a clear call to action: [E.G., LINK BACK TO FULL POST / ASK A QUESTION / INVITE REPLIES].

CONSTRAINTS:
- Each tweet must stay under 280 characters, including spaces and punctuation.
- Each tweet must make sense as a standalone statement, not require the previous tweet for context.
- Do not use hashtags unless explicitly requested.
- Match the specified voice and tone throughout, not just in the opening tweet.
- Number each tweet (1/, 2/, etc.) at the start.

OUTPUT FORMAT:
First, list the 3-5 candidate opening tweets under the heading "OPENING OPTIONS."
Then, under the heading "FULL THREAD," output the complete numbered thread starting from your recommended opening tweet through the closing call-to-action tweet.