I will write and engineer reusable AI prompts and a custom prompt library

I will design, write, and engineer reusable AI prompts plus a structured, custom prompt library your whole team can run on demand. Built for ChatGPT, Claude, and Gemini.

What you get

Most "prompt help" is a single clever paragraph that breaks the moment your inputs change. This service is different: I build reusable, parameterized prompt systems that hold up across real work, not one-off magic tricks.

• Engineered, reusable prompts with clearly marked variables (e.g. {audience}, {tone}, {product}, {constraints}) so the same prompt works across dozens of cases without rewriting.
• A custom prompt library delivered as an organized document (Google Doc, Notion, Markdown, or a clean spreadsheet) grouped by category, use case, and difficulty, with a table of contents and quick-start notes.
• Role / system prompts that set persona, voice, guardrails, and output format so your assistant behaves consistently every session.
• Few-shot examples baked into prompts where they materially improve quality, plus negative examples ("do not do X") where the model tends to drift.
• Structured output templates (Markdown, JSON, tables, or step lists) so downstream tools and teammates can actually parse the results.
• Chained / multi-step prompt flows for tasks too big for one shot, with handoff instructions between steps.
• A short usage guide: how to fill the variables, what to tweak, common failure modes, and how to recover when an answer goes sideways.
• Model notes covering how each prompt should be adjusted for ChatGPT (GPT-4 class), Claude, and Gemini, since they reward slightly different phrasing.
• Before/after samples showing a weak prompt and the engineered version, with the output difference, so you can see and trust the lift.

Everything is written in clear English, documented, and yours to keep, edit, and reuse forever.

Plans

Feature	Basic	Standard	Premium
Engineered, reusable prompts	3 prompts	10 prompts	25+ prompts
Organized prompt library document	Single category	Multi-category + index	Full taxonomy + index + tags
Role / system prompt included	1	2	Up to 5
Variables & parameterization	Basic	Yes	Advanced + presets
Structured output templates	—	Yes	Yes
Multi-step / chained flows	—	1 flow	Up to 3 flows
Model-specific notes (ChatGPT/Claude/Gemini)	1 model	2 models	3 models
Before/after sample outputs	—	Yes	Yes + tuning log
Usage & troubleshooting guide	Short	Yes	Extended
Revisions	1	2	3
Delivery format	Markdown / Doc	Doc / Notion / Sheet	Your choice + handoff call

Need something outside these tiers, like a domain-specific library (legal intake, e-commerce listings, support macros, cold outreach)? Message me first and I'll scope a custom package.

How it works

1. You brief me. Tell me your goal, the AI tool(s) you use, your audience, tone, and any examples of "good" output. A filled-in intake form or a few sample tasks is perfect.
2. I map the use cases. I break your goal into the recurring jobs the prompts must do, then group them into a library structure so nothing overlaps or gets lost.
3. I engineer the prompts. I draft each prompt with clear roles, variables, constraints, and output formats, then stress-test them against realistic inputs and iterate on the wording.
4. I document the library. I assemble everything into an organized, navigable document with a quick-start section and per-prompt notes.
5. You review. You run the prompts on your own tasks and flag anything that misses. I refine within your plan's revision count.
6. You receive the final pack. Clean files, ready to paste in, share with your team, and reuse indefinitely.

Why choose this

I treat prompts as small pieces of software, not lucky phrases. That means versioned wording, explicit inputs and outputs, edge-case handling, and documentation, so results stay stable when the person, the topic, or the day changes.

I optimize for handoff, not dependence. The deliverable is structured so a teammate who has never written a prompt can fill the variables and get consistent results. You should not need me again to keep using what I build, that is the point.

I also keep it honest. AI models are non-deterministic, so I won't promise identical output every run. What I will do is reduce variance, format outputs predictably, and give you the controls and fallbacks to steer quality yourself.

Who it's for / use cases

• Founders and small teams who want a repeatable content, support, or research workflow without hiring a prompt specialist.
• Marketers and writers building libraries for blog drafts, ad variations, email sequences, SEO briefs, and repurposing.
• Customer support / ops teams standardizing reply macros, summarization, and ticket triage.
• Agencies that want a white-label prompt pack to deploy across multiple clients.
• Developers wiring prompts into apps who need clean system prompts and reliable JSON/structured outputs.
• Course creators, coaches, and consultants packaging their methodology into guided AI workflows.

FAQ

Q: Which AI tools do these prompts work with? They're written primarily for ChatGPT (GPT-4 class), Claude, and Gemini, and I include notes on adapting phrasing per model. The core structure also transfers to most modern chat-based LLMs.

Q: Will the exact same output come out every time? No, and anyone who promises that is overselling. LLMs are probabilistic. I engineer the prompts to minimize drift, lock the output format, and include fallback instructions so quality stays high and predictable even when wording varies.

Q: What do you need from me to start? Your goal, the AI tool you use, your audience and tone, and ideally two or three real examples of tasks plus what "good" looks like. I'll send a short intake form to make this easy.

Q: Do I own the prompts and library? Yes. Everything I deliver is yours to keep, edit, reuse, and share with your team or clients. There are no usage limits from my side.

Q: Can you build a library for my specific industry or workflow? Absolutely, that's the ideal use case. Whether it's e-commerce listings, recruiting outreach, legal intake, or support replies, message me first so I can scope the right categories and depth.

Q: How are revisions handled? Each plan includes a set number of revision rounds. You test the prompts on real tasks, send me what misses, and I refine wording, variables, or structure within that count. Premium includes a tuning log so you can see what changed and why.

Q: Can you connect these prompts to my app or automation? I deliver clean system prompts and structured-output templates designed to drop into an app or automation, and I'll document the expected inputs/outputs. I don't build the integration code itself, but I make the prompt side integration-ready.

Q: How will I receive the final files? As an organized document in the format you prefer, Markdown, Google Doc, Notion, or a spreadsheet, with a table of contents, per-prompt notes, and a quick-start guide so you can use it immediately.