AI Agents & Prompt Templates for Manufacturing

You are a Manufacturing Flow SOP Generator — an AI documentation specialist that creates professional Standard Operating Procedures through guided discovery. Your method: Ask questions ONE AT A TIME. Each question uncovers more detail about the process. Tell the user to expect about 15 questions, but ask as many as needed. Complex procedures may require 20-30+. START by asking: "What procedure would you like to document? Give me the name of the process and a one-sentence description of what it accomplishes." Then ask: "Do you have any existing documentation — old SOPs, checklists, notes, training videos, or photos? If so, upload them and I'll use them as a starting point." DISCOVERY QUESTIONS (pick the most relevant next question based on what you've learned): Purpose & Scope: - What is the end goal of this procedure? - Who performs this task? What's their skill level? - How often is this done? - What triggers this procedure to start? Inputs & Outputs: - What materials/inputs are required? - What tools or equipment are needed? - What does a successful output look like? - How do you know when you're done? The Steps (most critical — probe deeper on each step): - "Walk me through the first thing you do." - "What do you do right before that?" - "How do you know when to move to the next step?" - "What could go wrong here? How would you know?" - "Are there any tricks or tips that aren't obvious?" - "What's the most common mistake people make?" Edge Cases & Safety: - What if something goes wrong at step X? - Are there safety considerations? - What should someone NEVER do? - Any environmental or cleanup requirements? RULES: - Ask ONE question at a time. Wait for the answer before asking the next. - Never assume you understand a step — always confirm. - Push for specifics: "How much?" "How long?" "How do you know?" - Flag safety-critical items with ⚠️ WARNING callouts. WHEN DONE, generate a complete SOP with: - Document header (title, author, date, revision, approval) - Purpose and scope - Safety requirements and PPE - Required tools, materials, and equipment - Numbered step-by-step procedure (with sub-steps) - Quality checkpoints - Troubleshooting guide - Sign-off section Footer: "Powered by ManufacturingFlow.ai | Generated [Today's Date]"

I'm interviewing an experienced [JOB TITLE] who has worked with [EQUIPMENT/PROCESS] for [X YEARS]. They are [RETIRING/TRANSFERRING] in [TIMEFRAME]. Generate a structured interview guide to capture their tribal knowledge. Focus on: 1. Process secrets — "What do you do differently that isn't in any manual?" 2. Failure patterns — "What failures have you seen and how did you fix them?" 3. Setup tricks — "What shortcuts or techniques make setup faster/better?" 4. Quality indicators — "How do you know something is about to go wrong?" 5. Workarounds — "What official procedures do you modify, and why?" 6. Undocumented dependencies — "What upstream/downstream factors affect this process?" Format as interview questions I can ask directly. Include follow-up probes for each category. Goal is to capture knowledge that exists nowhere except in this person's head.

I have rough notes from an interview with an experienced operator about [PROCESS/EQUIPMENT]. Convert these into a structured knowledge document. Here are my raw notes: [PASTE YOUR NOTES, TRANSCRIPT, OR BULLET POINTS HERE] Create a structured document with: 1. Process overview (what this covers and why it matters) 2. Key procedures (step-by-step, based on the expert's actual method) 3. Critical tips and tricks (the non-obvious knowledge) 4. Common failure modes and solutions 5. Quality indicators to watch for 6. Things NOT to do (common mistakes) Write in clear, direct language. Preserve the expert's specific terminology and measurements. Flag anything that seems safety-critical.

Help me document the optimal operating parameters for [EQUIPMENT NAME] running [PRODUCT NAME]. Current settings (as they know them): [LIST CURRENT PARAMETER SETTINGS: temperature, speed, pressure, etc.] I want to capture: 1. Nominal/target value for each parameter 2. Acceptable range (upper/lower limits) 3. How the operator knows if a parameter is drifting out of spec 4. What adjustments to make and in what order 5. Why each parameter matters (impact if too high/too low) 6. Dependencies between parameters (which settings affect each other) 7. Environmental factors that affect settings (temperature, humidity, material variation) 8. How long to run before checking if parameters are stable Format as a parameter reference card that can be posted by the machine and in the digital system.

Create a shift lead handoff document template based on this scenario: Shift: [DAY/SWING/NIGHT] Time frame: [DURATION OF SHIFT] The shift lead needs to communicate to the incoming shift lead: 1. Current equipment status (what's running, what's down, why) 2. Material status (what's in, what's expected, any issues) 3. Quality issues identified during this shift 4. Staffing or personnel notes relevant to operations 5. Upcoming activities (maintenance windows, changeovers, special orders) 6. Things the incoming shift should watch for or be aware of 7. Ongoing problems that aren't fully resolved 8. Positive highlights (good performance, milestones) Format as a one-page handoff form that takes 5-10 minutes to complete but ensures nothing critical falls through the cracks between shifts.

Create a training plan for a new [JOB TITLE] learning to operate [EQUIPMENT NAME]. Training duration: [X DAYS/WEEKS] Trainer: [EXPERIENCED OPERATOR / SUPERVISOR] Prerequisites: [ANY REQUIRED CERTIFICATIONS OR EXPERIENCE] Include: 1. Day-by-day training schedule 2. Skills to demonstrate at each stage 3. Hands-on exercises with increasing complexity 4. Safety training checkpoints 5. Quality standards they must meet 6. Assessment criteria for sign-off 7. Common mistakes new operators make and how to avoid them The trainee should be production-ready by end of training with documented competency verification.

Create a knowledge check quiz for operators who completed training on [EQUIPMENT/PROCESS]. Include 15 questions: - 5 safety questions (must get 100%) - 5 procedure questions (must get 80%) - 5 troubleshooting scenario questions (must get 80%) Mix question types: - Multiple choice - True/false - Short scenario-based ("What would you do if...") Include an answer key with explanations for each correct answer. Flag which questions are safety-critical (automatic fail if wrong). Format as a printable quiz with space for operator name, date, trainer signature, and score.

Create a safety training module for [EQUIPMENT NAME] suitable for [BEGINNER/ALL OPERATORS]. Safety topics to cover: [LIST KEY HAZARDS: electrical, mechanical, chemical, ergonomic, etc.] The module should include: 1. Hazard identification (what can go wrong and why) 2. Personal protective equipment (PPE) requirements 3. Lockout/tagout procedures (if applicable) 4. Emergency procedures (what to do if something goes wrong) 5. Safe work practices (do's and don'ts) 6. Common near-misses and accident scenarios 7. Reporting requirements for injuries/incidents 8. How to get help or ask questions Format as training material that can be delivered in 15-30 minutes. Include visuals descriptions (where diagrams would go). Add a sign-off section documenting that the operator completed and understands this training.

Create a cross-training module for maintenance technicians on [EQUIPMENT NAME/SYSTEM]. Target audience: Experienced maintenance technicians with [X YEARS] experience, but new to this specific equipment. Training should cover: 1. Equipment overview and basic operation 2. Common preventive maintenance tasks and frequencies 3. Diagnosis procedures for common failures 4. Disassembly/reassembly procedures for major components 5. Replacement parts inventory and part numbers 6. Lubrication points and specifications 7. Calibration requirements 8. When to escalate to OEM support or engineering 9. Safety considerations specific to this equipment 10. Documentation and record-keeping requirements Format as a practical guide that technicians can reference while working. Include decision trees for common repair scenarios.

Analyze this production data and provide insights: [PASTE YOUR DATA — CSV, table, or description of metrics] I want to understand: 1. Overall production efficiency (OEE if possible) 2. Top 3 causes of downtime or scrap 3. Trends over time (improving or declining?) 4. Bottleneck identification 5. Shift-to-shift or line-to-line comparisons 6. Specific actionable recommendations Present findings as: - Executive summary (3 bullet points) - Detailed analysis with numbers - Recommended actions ranked by impact - Suggested KPIs to track going forward

Create a shift handover report from these notes: Shift: [DAY/SWING/NIGHT] Date: [DATE] Operator: [NAME] Raw notes: [PASTE YOUR END-OF-SHIFT NOTES HERE] Format as a structured report with: 1. Production summary (units produced vs. target) 2. Quality issues (scrap count, rework, holds) 3. Equipment status (any issues, maintenance needed) 4. Safety incidents or near-misses 5. Material status (shortages, incoming) 6. Open items for next shift (specific actions needed) 7. Anything unusual the next shift needs to know Keep it concise — the incoming shift should be able to read this in 2 minutes.

Create a weekly production summary report from this data: Week of: [DATE RANGE] Location/Line: [FACILITY OR PRODUCTION LINE] Production metrics: [PROVIDE: total units produced, units targeted, on-time delivery %, scrap rate, downtime hours, safety incidents] Customer orders shipped: [NUMBERS] Orders on hold: [REASONS IF ANY] Key challenges: [LIST TOP ISSUES THIS WEEK] The report should include: 1. Production vs. target (achieve %, YTD trend) 2. Quality summary (defect rate, major issues) 3. Downtime analysis (equipment failures, causes) 4. Safety summary (incidents, near-misses, corrective actions) 5. Staffing/attendance notes 6. Upcoming week outlook 7. Key action items and who owns them Format as a one-page executive summary suitable for management/stakeholder distribution.

Help me analyze Overall Equipment Effectiveness (OEE) for [EQUIPMENT/LINE]. Current performance data: - Scheduled uptime: [X HOURS] - Actual uptime: [Y HOURS] - Planned production rate: [UNITS/HOUR] - Actual production rate: [UNITS/HOUR] - Good parts produced: [NUMBER] - Total parts produced: [NUMBER] Calculate: 1. Availability % (uptime / scheduled time) 2. Performance % (actual rate / planned rate) 3. Quality % (good parts / total parts) 4. Overall OEE % (Availability × Performance × Quality) 5. Which factor (Availability, Performance, Quality) is the biggest drag on OEE? For the weakest factor, provide: - Root causes - Improvement opportunities (quick wins vs. long-term) - Estimated impact if improved - Recommended actions with timeline Target: What would OEE be if we improved each factor by 10%?