AI Prompt Library

Write a comprehensive, SEO-optimized blog post based on this outline: [OUTLINE]. Include a compelling introduction, structured H2/H3 headers, actionable takeaways, and a conclusion with a call to action. Target audience: [AUDIENCE]. Tone: [TONE].

Write an engaging email newsletter about [TOPIC]. Include: subject line, preview text, hook opening, 3 key sections, and a single clear CTA. Keep it under 400 words. Tone: conversational and friendly.

Write a professional press release for [ANNOUNCEMENT]. Include headline, dateline, lead paragraph (who, what, when, where, why), quotes from [SPOKESPERSON], boilerplate, and contact info. Follow AP style.

Write a persuasive product description for [PRODUCT NAME]. Highlight: key benefits (not features), target customer pain points, and unique differentiators. Include a power headline, 3-4 benefit bullets, and a closing urgency statement. Under 150 words.

Write a thought leadership LinkedIn article about [TOPIC] from the perspective of a [JOB TITLE]. Include personal story/hook, 3 key insights with examples, practical takeaways, and end with a discussion question. Target: 800-1200 words.

Write a compelling cover letter for a [JOB TITLE] position at [COMPANY]. My background: [YOUR BACKGROUND]. Emphasize how my [SKILL 1] and [SKILL 2] directly address their needs. Show company knowledge, avoid clichés, keep to 3 paragraphs.

Write a crisp executive summary for this document: [PASTE DOCUMENT]. Limit to 1 page. Include: situation overview, key findings (3-5 bullets), recommended actions, expected outcomes, and timeline. Audience: C-suite executives.

Write a 3-email cold outreach sequence for [PRODUCT/SERVICE] targeting [PERSONA]. Email 1: Value-led hook, no pitch. Email 2 (3 days later): Specific pain point + social proof. Email 3 (5 days later): Soft close + easy reply. Each email under 100 words.

Review this code for: bugs, security vulnerabilities, performance issues, code smells, and best practice violations. For each issue found, explain: (1) the problem, (2) why it matters, (3) suggested fix with code example. ``` [PASTE CODE] ```

Write comprehensive unit tests for this function using [JEST/PYTEST/etc]. Cover: happy path, edge cases, error conditions, and boundary values. Use descriptive test names in the format "should [behavior] when [condition]". ``` [PASTE FUNCTION] ```

Refactor this legacy code to be clean, maintainable, and modern. Apply: meaningful variable names, single responsibility principle, DRY, proper error handling, and [LANGUAGE] best practices. Explain each change. ``` [PASTE CODE] ```

Design a RESTful API for [FEATURE/SYSTEM]. Include: endpoint list (method + path + description), request/response schemas with examples, authentication approach, error response format, and rate limiting strategy. Follow REST best practices.

Write an optimized SQL query to [WHAT YOU WANT TO DO] using the following schema: [TABLE SCHEMAS] Requirements: [SPECIFIC REQUIREMENTS]. Include indexes that would improve performance and explain your JOIN choices.

I'm getting this error: ``` [ERROR MESSAGE + STACK TRACE] ``` In this code: ``` [RELEVANT CODE] ``` Explain: (1) what's causing the error, (2) why it's happening, (3) the exact fix, (4) how to prevent it in the future.

Design a scalable system for [SYSTEM DESCRIPTION] that handles [SCALE: e.g., 1M users, 10k req/sec]. Cover: high-level architecture diagram (text), key components, data storage decisions, caching strategy, API design, and how you'd handle failures.

Write a regex pattern to match [WHAT TO MATCH] in [LANGUAGE]. Examples that should match: [EXAMPLES]. Examples that should NOT match: [COUNTER-EXAMPLES]. Explain each part of the pattern and provide test code.

Conduct a market research analysis for [PRODUCT/SERVICE] in [INDUSTRY]. Cover: market size & growth, target segments, top 5 competitors (strengths/weaknesses), customer pain points, pricing landscape, and key trends. Format as a structured report.

Create a competitive analysis comparing [YOUR PRODUCT] to [COMPETITOR 1], [COMPETITOR 2], and [COMPETITOR 3]. For each, evaluate: pricing, features, target market, positioning, strengths, weaknesses. Conclude with differentiation opportunities.

Summarize the current state of research on [TOPIC]. Cover: key findings from major studies, points of consensus, ongoing debates, gaps in the literature, and implications for practice. Cite sources where known. Format for academic audience.

Perform a detailed SWOT analysis for [COMPANY/PRODUCT/PROJECT]. For each quadrant, provide 5+ specific, evidence-based points rather than generic observations. Conclude with 3 strategic recommendations based on the analysis.

Create 3 detailed customer personas for [PRODUCT/SERVICE]. For each persona include: name, demographics, job role, goals, frustrations, typical day, how they find solutions, objections to purchase, and what messaging resonates. Make them feel like real people.

Analyze the top 5 emerging trends in [INDUSTRY] for [YEAR]. For each trend: explain the driving forces, current adoption level, impact timeline (1/3/5 years), affected business models, and opportunities for early movers.

Write a concise business plan for [BUSINESS IDEA]. Include: executive summary, problem/solution, target market (with size), revenue model, go-to-market strategy, competitive moat, team (roles needed), key metrics, and funding requirements. Keep each section punchy and investor-ready.

Write a compelling 10-slide pitch deck script for [STARTUP]. Follow the structure: Problem → Solution → Why Now → Market Size → Product → Business Model → Traction → Team → Competition → Ask. Each slide: headline + 3 talking points + key stat.

Create Q[QUARTER] OKRs for a [COMPANY TYPE] team focused on [AREA]. Write 3 Objectives, each with 3-4 measurable Key Results. OKRs should be ambitious (70% achievement = success), time-bound, and aligned to [COMPANY GOAL].

Create a structured agenda for a [MEETING TYPE] meeting (duration: [TIME]). Include: objectives, pre-read materials needed, timed agenda items with owners, decision points, parking lot section, and next steps format. Make it action-oriented, not update-based.

Write a project proposal for [PROJECT NAME]. Include: problem statement, proposed solution, scope & deliverables, timeline with milestones, resource requirements, budget estimate, risk assessment, and success metrics. Target audience: [STAKEHOLDER].

Write a balanced, constructive performance review for an employee who [DESCRIBE PERFORMANCE]. Include: 3 specific strengths with examples, 2 development areas with actionable suggestions, rating justification, and SMART goals for next period. Tone: professional and supportive.

Create 5 compelling short story premises in the [GENRE] genre. Each premise should include: a one-line hook, the central conflict, protagonist description, unique setting detail, and the core emotional theme. Make them feel original, not clichéd.

Develop a complex, memorable character named [NAME] for a [GENRE] story. Include: physical description, backstory, core motivation, fatal flaw, speech patterns, contradictions that make them human, relationship to the theme of [THEME], and character arc.

Write a vivid scene where [CHARACTER] confronts [SITUATION]. Show, don't tell. Use: sensory details, subtext in dialogue, character-specific voice, internal thoughts (limited), and tension building. Genre: [GENRE]. Mood: [MOOD]. 500-700 words.

Build the world of [WORLD NAME] for a [GENRE] story. Define: geography & climate, social structure & power systems, technology/magic rules, economy & currency, history (key events), cultural traditions, and current conflict driving the story.

Write a [POEM STYLE: haiku/sonnet/free verse/villanelle] about [TOPIC/EMOTION]. Avoid clichés. Use: concrete imagery, unexpected metaphors, sonic texture (alliteration, assonance), and leave space for the reader to feel rather than being told what to feel.

Explain [COMPLEX TOPIC] to someone with no background in the subject. Use: simple analogies to everyday things, avoid jargon (or define it immediately), a step-by-step build-up of concepts, and a concrete real-world example. End with a quick check: "Now you should understand..."

Create a detailed study plan to learn [SKILL/SUBJECT] in [TIMEFRAME]. I'm starting from: [CURRENT LEVEL]. Include: weekly breakdown, specific resources (books/courses/practice), daily time commitment, milestone checkpoints, and how to measure progress.

Create a 10-question quiz on [TOPIC] at [BEGINNER/INTERMEDIATE/ADVANCED] level. For each question: write the question, 4 multiple choice options (one correct, three plausible distractors), the correct answer, and a one-sentence explanation of why it's correct.

Create a [DURATION] lesson plan for teaching [TOPIC] to [AUDIENCE: age/level]. Include: learning objectives (measurable), hook/engagement activity, main instruction approach, hands-on activity or practice, formative assessment, and closure. Include time estimates.

Create a concept map for [TOPIC] showing relationships between key ideas. Use text-based format: main concept in center, branch to 5-7 sub-concepts, each with 2-3 connecting ideas. Show how concepts relate with labeled connections (e.g., "leads to", "is a type of").

Create a data analysis plan for this business question: [QUESTION]. Include: hypothesis, data sources needed, key metrics to calculate, statistical methods to use, visualization recommendations, potential confounders to control for, and how to present findings to [AUDIENCE].

Design a metrics dashboard for [ROLE: e.g., Marketing Director] tracking [BUSINESS GOAL]. Specify: 5-8 KPIs with formulas, visualization type for each (chart/table/number), data refresh frequency, drill-down capabilities, and alert thresholds.

Write a [Excel/Google Sheets] formula to [WHAT YOU WANT TO DO]. My data: [DESCRIBE STRUCTURE]. The formula should handle: [EDGE CASES]. Explain how the formula works step by step so I can modify it.

Design a survey to measure [WHAT YOU WANT TO MEASURE] from [AUDIENCE]. Include: 10-15 questions mixing: closed-ended (Likert scale, multiple choice), one open-ended, screener at start. Sequence logically, avoid leading questions, estimate completion time.

Create a 2-week social media content calendar for [BRAND] across [PLATFORMS]. For each day: platform, content type (video/image/text), topic, hook/caption, hashtags, and posting time. Mix: 40% educational, 30% entertainment, 20% promotional, 10% community.

Write 5 variations of ad copy for [PRODUCT/SERVICE] targeting [AUDIENCE]. For each variation use a different hook: (1) Fear of missing out, (2) Social proof, (3) Problem/solution, (4) Curiosity, (5) Direct benefit. Include headline, body (50 words), and CTA.

Write high-converting landing page copy for [PRODUCT/SERVICE]. Structure: hero headline (outcome-focused), subheadline, 3 benefit sections (each: mini-headline + description + icon idea), social proof section, FAQ (5 objections), and 3 CTA variations. Target: [AUDIENCE].

Create an SEO content brief for an article targeting the keyword "[KEYWORD]". Include: search intent analysis, target word count, title tag options (3), meta description (2 options), H1, content outline with H2/H3 structure, internal link opportunities, and semantic keywords to include.

Develop a brand voice guide for [COMPANY]. Define: 3 core personality traits with do/don't examples, tone variations (formal vs casual contexts), vocabulary to use/avoid, example rewrites of generic text in brand voice, and emoji/punctuation guidelines.

Help me prioritize this task list using the Eisenhower Matrix: [PASTE YOUR TASK LIST] For each task: assign quadrant (Urgent+Important / Important / Urgent / Neither), estimated time, and recommended action. Then give me a prioritized daily schedule.

I need to decide between [OPTION A] and [OPTION B]. Decision factors: [CRITERIA]. Help me by: (1) scoring each option 1-10 on each criterion, (2) identifying risks/downsides I may not have considered, (3) recommending the choice with reasoning, (4) suggesting a reversibility check.

Write a Standard Operating Procedure (SOP) for [PROCESS]. Include: purpose & scope, roles & responsibilities, prerequisites, step-by-step instructions (numbered, specific), decision points with if/then branches, quality checkpoints, and common troubleshooting. Format for [AUDIENCE].

Review the UX copy in this UI for clarity, tone, and conversion. For each text element: rate it 1-5 for clarity, identify the problem if below 4, and provide an improved version. Focus on: buttons, error messages, empty states, onboarding, and confirmation dialogs. [PASTE UI TEXT LIST]

Create a cohesive color palette for a [BRAND TYPE] brand with personality: [ADJECTIVES]. Provide: primary color (hex), 2 secondary colors, 2 neutral shades, 1 accent color, semantic colors (success/warning/error/info). For each: usage context and psychological effect.

Write a design brief for [PROJECT: e.g., mobile app redesign]. Include: project overview, problem statement, target users (with goals & frustrations), scope of work, design constraints, success metrics, competitors to study, and deliverables. Format for freelance designer.

Write a production-grade system prompt for an AI assistant that [ROLE/PURPOSE]. The assistant must: [KEY BEHAVIOURS]. It should never: [RESTRICTIONS]. Tone: [TONE]. Include: role definition, task scope, output format instructions, fallback behaviour when uncertain, and a worked example of ideal input/output.

Design an AI agent for [TASK]. Define: (1) the trigger/input that starts the agent, (2) the tools/APIs it needs access to, (3) the decision tree it follows, (4) how it handles ambiguity or failure, (5) the output format, (6) guardrails to prevent errors. Make it production-ready.

Design a Retrieval-Augmented Generation (RAG) pipeline for [USE CASE] using [DOCUMENT TYPES]. Cover: chunking strategy (size, overlap, method), embedding model choice and why, vector store selection, retrieval approach (semantic vs hybrid), re-ranking, prompt template, and how to handle no-match cases.

Build a multi-step prompt chain to accomplish: [GOAL]. Break it into sequential steps where each LLM call's output feeds the next. For each step: input format, prompt template, expected output format, and validation check. Include error handling if a step produces unexpected output.

Create an evaluation rubric to score AI-generated [OUTPUT TYPE: e.g., summaries, emails, code]. Define 5 criteria, each scored 1-5. For each criterion: name, what "5" looks like vs "1", and example good/bad outputs. Include an overall scoring formula and threshold for "acceptable" quality.

Generate 20 high-quality training examples for fine-tuning an LLM to [TASK]. Format as JSONL with "prompt" and "completion" fields. Examples should: cover diverse phrasings of the same task, include edge cases, show the exact tone/style/format you want, and avoid any examples the model might hallucinate on.

Write a Model Context Protocol (MCP) tool specification for a tool that [TOOL PURPOSE]. Include: tool name, description (written for an LLM to understand when to use it), input schema (JSON Schema with types, required fields, descriptions), output schema, example call, and error cases the LLM should handle.

Red-team this AI system prompt for safety and robustness: [PASTE SYSTEM PROMPT] Test for: (1) prompt injection vulnerabilities, (2) jailbreak paths, (3) information leakage, (4) off-topic behaviour, (5) adversarial user inputs. For each vulnerability found: describe the attack, show a sample exploit, and suggest a mitigation.

Describe an agentic workflow for [BUSINESS PROCESS] as a text-based flowchart. Show: trigger → agent steps → tool calls → decision branches → outputs. Use ASCII boxes and arrows. Mark which steps are automated vs require human approval. Note: latency-sensitive steps that should run in parallel.

Write a cold email to [PROSPECT ROLE] at [COMPANY TYPE]. Lead with their pain: [SPECIFIC PAIN POINT]. Do NOT mention our product in the first sentence. Structure: hook (their world, not yours) → why this matters now → one-line solution → single low-friction CTA. Under 80 words. No subject line fluff.

Write a 4-touch follow-up sequence for [PROSPECT] who didn't reply to my first email about [OFFER]. Touch 2 (day 3): different angle, shorter. Touch 3 (day 7): social proof or case study. Touch 4 (day 14): break-up email with easy reply option. Each under 60 words. No guilt-tripping.

Write a LinkedIn connection request note for [PROSPECT NAME], [THEIR ROLE] at [COMPANY]. My reason for connecting: [REAL REASON]. Max 280 characters. Sound like a human, not a sales bot. Reference something specific — their post, company news, or shared background. No pitch in the note.

I'm selling [PRODUCT/SERVICE] and just heard this objection: "[OBJECTION]". Write 3 different responses — one that validates and pivots, one that reframes the objection, and one that uses a question to uncover the real concern. Keep each under 50 words. No dismissing or bulldozing.

Write a discovery call script for selling [PRODUCT] to [BUYER PERSONA]. Include: rapport opener (30 sec), agenda-setting, 8 diagnostic questions (ordered from safe→deep), a section to uncover budget/authority/timeline, and a clean next-step close. Mark which questions are mandatory vs skip-if-time.

Write a compelling case study for [CUSTOMER] who used [PRODUCT] to solve [PROBLEM]. Structure: headline with result metric, customer background (2 sentences), the challenge they faced, why they chose us, implementation (brief), results with specific numbers, and a quote from [CUSTOMER CONTACT]. 400 words max.

Write a sales proposal for [PROSPECT COMPANY] for [SOLUTION]. Sections: executive summary (their words, their pain), proposed solution (tailored, not generic), implementation timeline, investment summary (clear pricing, no surprises), ROI calculation (conservative), social proof, next steps with specific date. Keep it scannable — they won't read walls of text.

Analyse why we [WON/LOST] the deal with [COMPANY]. What I know: [DEAL DETAILS]. Help me identify: (1) the real buying criteria vs stated criteria, (2) where our pitch missed, (3) what the competitor did better (or what we did right), (4) process changes to improve next time. Be brutally honest.

I want to set up GiiS to search across [YOUR TOOLS: e.g., Notion, Slack]. Help me write a connector setup plan: (1) which connectors to prioritise based on where my team actually stores information, (2) indexing schedule (how often to re-sync), (3) which document sets to create, (4) test questions to verify the RAG is working correctly after setup.

Help me design a GiiS AI agent (persona) for [USE CASE: e.g., customer support, sales research, code review]. Define: (1) system prompt for the agent, (2) which tools to enable (web search, code interpreter, connectors), (3) what document sets it should have access to, (4) response format it should use, (5) when it should escalate to a human.

I want to use GiiS's sales loop to automate outreach for [BUSINESS TYPE]. Help me configure: (1) ideal customer profile to target (industry, size, role), (2) keywords for the lead scraper, (3) cold email sequence copy (3 emails), (4) reply classification criteria (what counts as "interested"), (5) escalation trigger for human follow-up. My product: [PRODUCT].

Design a developer workflow using c0d3r, GiiS's local coding agent. I work on: [TECH STACK]. Help me set up: (1) which built-in skills are relevant to my stack, (2) whether to connect Claude Code or Codex as a chat model via the desktop app, (3) a prompt for the agent to understand my codebase, (4) how to pipe GiiS search results into my coding agent.

I want to set up GiiS with my own AI model instead of a default one. Help me plan: (1) whether a cloud provider (GPT, Claude, Gemini, Bedrock, Azure OpenAI) or a local model via Ollama/LM Studio fits my needs, (2) what hardware I need if I go local, (3) how to configure the connection in GiiS, (4) how to switch models per-agent without breaking existing workflows.

I want to deploy GiiS for a team of [SIZE] people to keep all AI usage private (no data training). Help me plan: (1) which tier/plan fits (Free / Pro / Team / Enterprise), (2) how to set up shared document connectors vs personal connectors, (3) how to create role-based agents (different agents for different teams), (4) GDPR compliance checklist for our use.

I'm currently using ChatGPT for [USE CASES]. Help me migrate to GiiS: (1) map each ChatGPT use case to the equivalent GiiS feature, (2) identify anything GiiS does better, (3) list what I might miss and workarounds, (4) write onboarding instructions I can share with my team, (5) key settings to configure on day 1.

Write a reusable prompt template I can save in GiiS for [RECURRING TASK: e.g., summarising meeting notes, reviewing contracts, writing weekly reports]. The template should: use [BRACKETS] for variable parts, include format instructions for the output, specify the ideal response length, and handle the most common edge cases I'll encounter.