This free library contains 40 curated AI prompt templates for engineering tasks. Browse by category or search by keyword, then copy any prompt to use in ChatGPT, Claude, or any other LLM. Replace the {{variable}} placeholders with your project details and submit. Each prompt includes an example output preview so you know what to expect before you use it.
Code Generation — Python scripts for engineering calculations: NEC 220.82 residential load calcs, voltage drop (K-factor method), simplified Manual J HVAC loads, AISC 360 steel beam sizing, sensor data parsing and plotting, and API data fetching with error handling.
Spec Drafting — CSI MasterFormat specification sections: Division 26 electrical, Division 28 fire alarm (NFPA 72), mechanical equipment (ASHRAE 90.1), structured cabling (TIA-568-C), solar PV (Division 48 / NEC 690), and lighting design narratives.
Drawing Review — Compliance checking prompts: NEC 2023 electrical drawing review, IBC 2021 structural review, NFPA 72 2022 fire alarm riser review, ASHRAE 62.1 HVAC duct layout, and submittal review checklist generation.
RFI Writing — Construction communication: RFI response drafts, clarification requests, code conflict RFIs, RFI log summaries, and design clarification emails.
Calc Docs — Calculation documentation: professional headers, assumptions/limitations sections, code reference tables, results summary tables, and PE certification statements.
Report Writing — Field and technical reports: site inspection reports, commissioning test results, ASHRAE Level 2 energy audit summaries, construction observation reports, and arc flash study summaries.
Troubleshooting — Fault diagnosis: electrical fault analysis, HVAC fault tree, enterprise network troubleshooting, PLC/SCADA alarm sequence analysis, and structural crack assessment.
1. Browse by category using the filter tabs, or search by keyword to find the prompt you need. 2. Click the prompt card to read the full template. Variables are highlighted in amber — these are the parts you replace. 3. Click '📋 Copy' to copy the prompt text to your clipboard. 4. Paste into ChatGPT (chat.openai.com), Claude (claude.ai), Gemini, or any LLM interface. 5. Replace all {{variable}} placeholders with your project-specific information. 6. Add any additional context your project requires (project number, owner standards, output format preference). 7. Click '▼ Example Output' on any card to see what kind of output to expect.
Tip: For the best results, use GPT-4o or Claude 3.5 Sonnet. Specify the output format you need (e.g., 'format as a Word table', 'output Python code only', 'use SI units').
GPT-4o — Strongest for code generation and structured JSON output. Good for Python calculation scripts, API integrations, and data processing tasks. Available at chat.openai.com or via the OpenAI API.
Claude 3.5 Sonnet — Strongest for long-form structured writing and following complex formatting instructions. Ideal for spec drafting, report writing, and reviewing long documents. 200K context window handles full calc packages. Available at claude.ai.
Gemini 1.5 Pro — Good for tasks that involve Google Workspace integration (Docs, Sheets). 1M token context window for very long documents.
GPT-4o mini / Claude 3.5 Haiku — Cost-effective options for high-volume, simpler tasks like RFI log summarization or routine drawing review comments. Much cheaper per token than the flagship models.
All 40 prompts are starting points designed to be customized. Here are common additions that improve output quality:
Add output format: 'Format the output as a Word table with columns [...]' or 'Output Python code only with no explanation.'
Add code standards: 'Apply NEC 2023 (not NEC 2017)' or 'Reference AISC 360-22 (not AISC 360-16).'
Add firm standards: 'Our office standard uses CSI 3-part format and our spec boilerplate starts with...' Paste your standard template for the LLM to follow.
Add project constraints: 'This is a healthcare facility (NFPA 99)' or 'The owner specification requires 3 competitive bids per section.'
Add context for code generation: 'I am using Python 3.11 with numpy 1.26, pandas 2.1, and matplotlib 3.8. Generate type hints and a docstring.'
40 curated prompts across 8 categories: Code Generation (Python scripts for NEC load calcs, voltage drop, HVAC loads, structural beam sizing), Spec Drafting (CSI MasterFormat Division 26 electrical, fire alarm, mechanical, structured cabling, solar PV), Drawing Review (NEC, IBC, NFPA 72, ASHRAE), RFI Writing (response drafts, clarification requests, code conflict RFIs), Calculation Documentation, Report Writing, and Troubleshooting.
Copy the prompt template using the Copy button, paste it into ChatGPT or Claude, replace the {{variable}} placeholders with your project information, and submit. For best results use GPT-4o or Claude 3.5 Sonnet, and add your specific output format requirements.
Yes — all prompts are designed to be modified. Replace the {{variable}} placeholders with your project-specific details, and add your own output format, code standard versions, and project constraints. The templates are starting points, not rigid scripts.
GPT-4o and Claude 3.5 Sonnet are the strongest for most engineering tasks. Claude excels at structured writing and has a 200K context window ideal for reviewing long calc packages. GPT-4o excels at code generation. For high-volume simpler tasks, GPT-4o mini or Claude 3.5 Haiku are cost-effective.
Use the Code Generation category prompts. Copy a prompt, replace the variables, and paste into GPT-4o or Claude. Specify your Python version, available libraries (numpy, pandas, matplotlib), and ask for type hints and docstrings for production-quality code.