Grok 4.7 Prompt Generator

You are a senior software engineer agent. I am providing a product spec for a new feature: a CSV-to-dashboard pipeline for a SaaS analytics product. Spec: - User uploads a CSV (any schema) - System auto-detects column types (numeric, categorical, datetime) - Generates a live interactive dashboard (bar, line, pie charts) - User can filter by any column and export filtered data Deliver end-to-end: 1. Data model (TypeScript interfaces for parsed CSV + chart config) 2. CSV parsing service (Node.js) with type inference and error handling 3. Chart config generator that maps column types to recommended chart types 4. React dashboard component using Recharts — all chart types, filter sidebar, export button 5. Unit tests for the parsing service covering: empty CSV, mismatched column counts, mixed-type columns Architect first (2 paragraphs). Then produce all code. No placeholder sections.

I am the Chief Integration Officer leading the acquisition of a 200-person B2B SaaS company by a 1,200-person enterprise software business. Day 1 is in 6 weeks. Produce a complete 90-day integration playbook: DAY 1 (Week 1): - Communications: internal announcement script for both companies, external customer communication, FAQs for acquired company employees - IT: access provisioning checklist, email migration timeline, security credential audit - HR: benefits transition timeline, org chart consolidation plan, manager briefing guide DAY 30 (Weeks 2–4): - Product: feature parity gap analysis framework, roadmap consolidation process - Sales: CRM unification plan, cross-sell opportunity identification, territory conflict resolution - Finance: P&L consolidation model, cost synergy identification checklist DAY 90 (Weeks 5–13): - Culture: integration health survey design, team-building programme outline - KPIs: 10 metrics to track integration success, with targets and owners - Risk register: top 8 integration risks with likelihood, impact, and mitigation Be specific. Name owners as roles (CIO, VP Sales, etc.), not individuals.

Prove the following claim, then self-audit your proof before presenting it: Claim: For any prime p ≥ 5, p² ≡ 1 (mod 24). Work in two phases: PHASE 1 — PROOF: 1. State all definitions and assumptions 2. Prove the claim step by step using modular arithmetic 3. Verify the result holds for p = 5, 7, 11, 13 explicitly PHASE 2 — SELF-AUDIT: 1. Check each step for logical gaps or unjustified assumptions 2. Identify any step where a reader might object 3. Strengthen the weakest step 4. Confirm the proof is complete and correct, or revise it Present the final verified proof only after completing both phases. Label every step.

I am providing a 380,000-token Python monolith codebase that was built over 7 years and must be decomposed into microservices. Produce a complete decomposition plan: 1. SERVICE MAP — identify the natural service boundaries in the codebase. For each proposed service: name, responsibility, estimated LOC, external dependencies 2. DEPENDENCY GRAPH — which services call which, in ASCII diagram 3. DATA OWNERSHIP — which service owns which tables; identify shared tables that need splitting 4. MIGRATION SEQUENCE — the safest order to extract services, with rationale for the sequence 5. STRANGLER FIG PLAN — how to run the monolith and each new service in parallel during migration without downtime 6. RISK REGISTER — top 5 technical risks with probability (H/M/L), impact (H/M/L), and mitigation 7. TIMELINE — realistic 18-month milestone plan with team size assumptions [Paste full codebase] Do not summarise — produce a plan detailed enough for a VP Engineering to approve.

You are a test engineering agent. I am providing a Python REST API with 22 endpoints (FastAPI, PostgreSQL backend). Generate a complete test suite: 1. UNIT TESTS (pytest): for each service-layer function, test the happy path plus at least 2 edge cases each 2. INTEGRATION TESTS: for each endpoint, test: valid request, missing required field, invalid auth token, database unavailable (mock), rate limit exceeded 3. CONTRACT TESTS (Pact): consumer contract for the 5 most externally-consumed endpoints 4. PERFORMANCE BASELINE: a Locust script that load-tests the 3 highest-traffic endpoints at 500 concurrent users for 60 seconds 5. CI PIPELINE: a GitHub Actions workflow that runs unit → integration → performance tests in sequence, fails the build if unit or integration fail, and posts a Slack summary Organise tests so they can run independently. Include fixtures, mocks, and factory functions. [Paste API code] Think through the test architecture before writing any test code.

Write a board-level strategy memo for a Series C SaaS company ($18M ARR, 140% NRR, B2B HR tech) facing the following situation: Context: - Two strategic paths are under active board discussion: (A) raise a $50M Series D to pursue enterprise upmarket and international expansion, or (B) pursue a strategic acquisition by one of three inbound acquirers at 8–10× ARR ($144–180M) - The CEO is asking the board to choose at the next meeting in 3 weeks The memo must: 1. FRAME the decision with the key question the board must actually answer (not the superficial one) 2. ANALYSE Path A: what must be true for the raise to succeed, the top 3 risks, and the likely outcome range (base / bull / bear) 3. ANALYSE Path B: which acquirer is strategically optimal and why, negotiation leverage points, and the dilution/payout analysis for the founding team at each valuation 4. RECOMMEND a path with clear reasoning — do not hedge 5. IDENTIFY the 3 questions the board must answer before voting Tone: board-ready. Concise. No filler. Every sentence should earn its place.

Audit the following Python web application for security vulnerabilities. Work in two explicit phases: PHASE 1 — INITIAL SCAN: Find all security issues you can identify. For each issue: severity (Critical/High/Medium/Low), CWE number, the specific code location, and the attack scenario. PHASE 2 — ADVERSARIAL RE-AUDIT: Now assume you are an attacker who has read Phase 1. What did the Phase 1 audit miss? What vulnerabilities could be chained together? What did the audit correctly identify but underestimate? Add any findings from this second pass. FINAL REPORT: - Combined, de-duplicated finding list, ranked by severity - For each: reproduction steps, CVSS score estimate, and recommended fix - An overall risk summary (1 paragraph) [Paste code] Do not skip the Phase 2 re-audit — it is the most valuable part.

You are a go-to-market agent. A B2B SaaS company is launching a new AI-powered contract analysis feature in 4 weeks. Build the complete launch sequence: WEEK 1 — FOUNDATION: - Define launch success metrics (3 primary, 2 secondary) - Draft the ICP (ideal customer profile) for this specific feature - Write the positioning statement using the Geoffrey Moore formula WEEK 2 — CONTENT: - 5-email nurture sequence for existing customers (subject lines + body copy) - Product Hunt launch copy (tagline, description, first comment) - 3 LinkedIn posts for the founding team (different angles: founder story, customer problem, demo) WEEK 3 — SALES ENABLEMENT: - 1-page battle card (our feature vs 3 named competitors) - 10 discovery questions reps should ask to qualify for this feature - Objection handling guide: top 5 objections + responses WEEK 4 — LAUNCH DAY: - Hour-by-hour launch day checklist - Crisis comms template for 3 scenarios: bug in demo, competitor counter-announcement, low signup rate All copy must be specific and ready to use — no placeholders.

I am a compliance officer at a US-based fintech company planning to expand into the EU and UK markets in Q1 2027. Produce a regulatory gap analysis: 1. REQUIREMENTS MATRIX — for each regulation (GDPR, PSD2, UK FCA Consumer Duty, DORA, EU AI Act for financial services AI), list: what applies to us, our current compliance status (compliant / partial / gap), and the specific gap 2. PRIORITY RANKING — which gaps carry the highest legal/financial risk if unfilled at launch 3. REMEDIATION ROADMAP — for each gap: the fix, the owner (role), estimated effort (S/M/L), and dependency on other fixes 4. EXTERNAL COUNSEL BRIEF — a 500-word brief to send to a UK/EU regulatory law firm, describing our business model, product, data flows, and the specific advice we need 5. BOARD SUMMARY — a 3-bullet executive summary a non-lawyer board member can understand Assume a 12-person engineering team and a 3-person compliance function. Be specific about what needs to change in our tech stack vs our policies vs our contracts.

Design and implement the core of a real-time collaborative text editor (like Google Docs): 1. ALGORITHM CHOICE: Explain why you choose OT (Operational Transformation) or CRDT for this use case. Prove mathematically that your chosen approach guarantees convergence for 3 concurrent users making conflicting edits at the same offset. 2. DATA STRUCTURES: Define the complete TypeScript types for: document state, operation, operation history, and client cursor state. 3. SERVER IMPLEMENTATION (Node.js/WebSocket): broadcast operations to all clients, maintain the canonical document state, handle late-joining clients who need to sync from any point in history. 4. CLIENT IMPLEMENTATION (React): a text editor component that sends operations on every keystroke, applies remote operations without cursor jump, and shows other users' cursors with names. 5. CONFLICT TEST: write a test that simulates user A inserting 'X' at position 5 and user B deleting characters 3–7 simultaneously. Show the system converges to the same document on both clients. Provide complete, working code. No pseudocode.

I am sharing a 3-statement financial model for a SaaS company. Perform a two-phase audit: PHASE 1 — FORMULA AUDIT: - Check all revenue formulas: do MRR → ARR → revenue calculations chain correctly? - Check cost structure: COGS, S&M, R&D, G&A as % of revenue — are the assumptions internally consistent? - Check cash flow: does the cash flow statement reconcile with net income and balance sheet changes? - Flag every formula that produces a surprising result PHASE 2 — ASSUMPTION STRESS TEST: - Identify the 5 assumptions the model is most sensitive to - For each: what happens to free cash flow if this assumption is 20% worse than the base case? - Which combination of bad assumptions produces a cash-out scenario, and when? FINAL VERDICT: - Is this model bankable? (yes / no / conditionally) - The 3 things that must be fixed before showing this to investors [Paste model or describe the key assumptions and formulas]

You are a developer experience agent. I am providing an existing documentation site for a developer API that has grown organically and is now inconsistent. Redesign and rewrite it: 1. AUDIT: list every structural problem with the current docs (missing sections, inconsistent terminology, no code examples, outdated endpoints) 2. INFORMATION ARCHITECTURE: design a new docs structure — table of contents with all pages and their purpose 3. GETTING STARTED (rewrite): new user can make their first authenticated API call in under 5 minutes. Python + JavaScript examples. 4. CORE CONCEPTS (rewrite): explain authentication, pagination, rate limits, and error handling each on their own page 5. API REFERENCE TEMPLATE: a reusable template for documenting each endpoint — method, path, description, parameters table, request example, response example, error codes 6. MIGRATION GUIDE: for the 3 endpoints that changed between v1 and v2 — before/after code showing exactly what to update [Paste existing documentation] Do not just critique — produce the rewritten content.

Build a vendor evaluation framework for selecting a new cloud data warehouse platform. Context: mid-market SaaS company, 500GB current data, 3× annual growth, 5-person data team, primary use cases are BI dashboards and ML feature stores. 1. REQUIREMENTS — define the non-negotiable requirements (must-have) vs nice-to-have, specific to our context 2. EVALUATION CRITERIA — score each criterion 1–5 with weight. Total must add to 100. Justify each weight. 3. VENDOR SHORTLIST — for each of 4 vendors (Snowflake, BigQuery, Redshift, Databricks): how each scores on your criteria, total weighted score 4. PROOF OF CONCEPT DESIGN — a 4-week PoC plan with specific tests that will expose the real differences between vendors for our use case 5. TOTAL COST OF OWNERSHIP — build a 3-year TCO model for each vendor with the key assumptions stated 6. RECOMMENDATION — which vendor to choose and the 3 deal terms to negotiate Produce the framework as a usable decision document, not advice.

Design and implement a zero-downtime migration strategy for the following change: Current schema: users table with a single name VARCHAR(255) column. Target schema: split into first_name VARCHAR(100) and last_name VARCHAR(100). Production database: PostgreSQL 16, 8M rows, live 24/7 with 500 writes/minute. Deliver: 1. MIGRATION PLAN: step-by-step sequence that allows the old and new schema to coexist during the migration window 2. BACKWARD-COMPATIBLE MIGRATION FILE (SQL): adds new columns without removing old ones, backfills data, adds constraints 3. APPLICATION CODE CHANGES (Python/SQLAlchemy): a dual-write layer that writes to both old and new columns during the transition 4. VERIFICATION QUERY: a SQL query that confirms the backfill is complete and consistent 5. CLEANUP MIGRATION: the final SQL to remove the old column once all application instances use the new schema 6. ROLLBACK PLAN: exact steps to reverse each phase if something goes wrong Identify the one step where data loss is possible and show how you prevent it.

Review the following system architecture for a real-time bidding (RTB) ad platform that must handle 500,000 bid requests/second with sub-10ms P99 latency. PHASE 1 — ANALYSIS: 1. Does the proposed architecture meet the latency and throughput requirements? Show your reasoning with rough calculations. 2. Identify the 3 most likely bottlenecks under peak load. 3. What single points of failure exist? PHASE 2 — ADVERSARIAL REVIEW: Now take the opposing position: argue that the architecture will fail. What are the weakest assumptions? What edge cases were not considered? PHASE 3 — SYNTHESIS: Reconcile both phases. What is your overall verdict? What are the top 3 changes that would have the highest impact on meeting the requirements? [Describe or paste architecture diagram] Label each phase clearly.

I am providing a 200-page government policy document on AI regulation in financial services (approximately 180,000 tokens). Produce an actionable analysis for a fintech General Counsel: 1. SCOPE: What companies, products, and activities does this regulation cover? What is explicitly excluded? 2. KEY OBLIGATIONS: The 10 most important compliance obligations, in plain language, ranked by compliance cost 3. DEFINITIONS: The 5 most important defined terms and how they differ from existing EU/UK regulatory vocabulary 4. TIMELINE: All compliance deadlines and what must be done by each 5. PENALTIES: Maximum penalties, how they are calculated, and the factors that increase or reduce them 6. GREY AREAS: The 3 provisions most likely to require regulatory guidance or litigation to interpret 7. ACTION PLAN: A 90-day compliance action plan for a 50-person fintech, with roles and workstream owners [Paste policy document] Do not paraphrase — extract, analyse, and prioritise.

You are a machine learning infrastructure engineer agent. Design and implement a production-grade model serving system: Requirements: - Serve a fine-tuned BERT model for text classification (20 categories) - P99 latency < 50ms at 1,000 requests/second - Automatic batching for GPU efficiency - Model versioning with zero-downtime blue/green deployment - Monitoring: latency, throughput, prediction confidence distribution, GPU utilisation Deliver: 1. SERVING ARCHITECTURE: diagram (ASCII) and component descriptions 2. TRITON INFERENCE SERVER CONFIG: model repository structure and config.pbtxt 3. BATCHING SERVICE (Python/FastAPI): dynamic batch accumulation with timeout, queue management, response routing 4. DEPLOYMENT SCRIPT (Docker + Kubernetes): Deployment, Service, HPA manifests 5. MONITORING SETUP (Prometheus + Grafana): the 5 most important metrics, alert rules, and dashboard JSON 6. LOAD TEST: a Locust script that ramps from 100 to 2,000 requests/second and reports P50/P95/P99 latency Think through the batching algorithm before writing any code — it is the most performance-critical component.

Build a crisis communications plan for a $500M ARR enterprise software company facing the following scenario: Scenario: A security researcher has publicly disclosed a critical vulnerability in our core product 12 hours before our planned patch release. 2,000 enterprise customers are affected. The vulnerability allows unauthenticated data access. Three customers have already reported suspicious activity. Produce: 1. FIRST 60 MINUTES — exact actions, by role (CEO, CTO, Head of Comms, General Counsel, Head of Customer Success), with draft copy for each customer-facing communication 2. CUSTOMER NOTIFICATION — tiered approach: affected customers with confirmed breach, affected customers with no breach detected, and all other customers. Draft each communication. 3. PUBLIC STATEMENT — a 400-word statement for the press and website that is accurate, non-alarmist, and demonstrates competence 4. SOCIAL MEDIA PROTOCOL — what to say, what not to say, how to handle public criticism on LinkedIn and X 5. REGULATORY NOTIFICATION — which regulators must be notified (GDPR, SEC, etc.), by when, and what must be included 6. 30-DAY RECOVERY PLAN — actions to rebuild customer trust post-incident Use real-world crisis communications best practices. Do not produce a generic template.

You are a competitive intelligence agent. Produce a full competitive landscape report for a B2B project management SaaS (targeting engineering teams at 50–500 person companies, $200 ACV, product-led growth). Work through this sequentially: 1. COMPETITOR MAP — categorise all competitors into: direct (same ICP, same price range), indirect (different ICP or price), and emerging (same ICP, venture-backed, < 3 years old) 2. FEATURE MATRIX — for the top 5 direct competitors: compare on 12 dimensions (key features, integrations, pricing model, free tier, mobile app, enterprise SSO, API, AI features, customer reviews average, funding, headcount trend) 3. POSITIONING ANALYSIS — for each direct competitor: their primary positioning claim, target buyer persona, and the gap in their positioning you could own 4. WIN/LOSS PATTERNS — based on the competitive matrix, what 3 situations does each competitor tend to win in? What 3 do they lose in? 5. STRATEGIC RECOMMENDATION — the 1 positioning move and 2 product investments that would most improve win rates against the top 3 competitors Be specific. Name real competitors. Do not produce a generic framework.

Critique the following business plan in two phases: PHASE 1 — INITIAL ANALYSIS: Evaluate: (1) Market size and validation, (2) Business model and unit economics, (3) Go-to-market strategy, (4) Team and execution risk, (5) Financial projections and assumptions. Rate each 1–5 and justify. PHASE 2 — STEELMAN THEN DESTROY: First, make the strongest possible case FOR this business (steelman). Then make the strongest possible case AGAINST it. Which case is more persuasive and why? FINAL VERDICT: - Would you invest? (yes / no / conditionally with what conditions) - The one question the founders must answer convincingly before raising - The one change to the plan that would most increase probability of success [Paste business plan] Do not hedge. Give a clear verdict with reasoning.