Engineering and architecture run on documents — specifications, design basis memos, review summaries, reports, RFIs, and tender responses. This pack gives you tested prompt templates for that technical writing, each with fill-in-the-blank variables so you start from a sound structure instead of a blank page.
How it works
Each prompt names your discipline, project type, and the deliverable, and asks the model
to produce the structure and narrative — never the engineering judgement. The bracketed variables —
[discipline], [project type], [deliverable] — shape the output to your project. Filter by
category or search, replace the brackets, copy, and paste into your AI tool.
Why verification is non-negotiable
Language models routinely cite the wrong code edition, fabricate clause numbers, and miss jurisdiction-specific requirements. These prompts deliberately separate the documentation (which AI handles well) from the analysis and code compliance (which it does not). Treat every standard reference, load, and value the model produces as a placeholder until you verify it against the actual standard.
Tips for better output
- Keep calculations in your validated tools. Use the prompts for the narrative around the numbers, and paste verified results in yourself.
- State the jurisdiction and code suite. Tell the model which standards govern (Eurocodes, ASCE, BS, IBC, etc.) and still confirm every reference it returns.
- Use the review and checklist prompts early. They catch coordination and constructability gaps cheaply at the concept and detailed-design stages.
- Protect project confidentiality. Describe the work generically and keep client names, site data, and proprietary detail out of public AI tools.