Browse categories
Explore
Fiverr Pro
English
$
USD
I will audit, debug, and improve your ai prompt
MdHarun
About this gig
Is your AI prompt giving inconsistent, vague, or unreliable results?
I will audit, debug, and improve your prompt so it is clearer, more structured, easier to test, and better aligned with your intended output.
My review can identify:
unclear or conflicting instructions
missing inputs or assumptions
weak output requirements
hallucination and grounding risks
formatting problems
missing failure-handling rules
unnecessary complexity or duplication
Depending on your package, I will provide an improved prompt, structured audit notes, and targeted test cases to help validate the revision.
My approach follows a simple principle:
Clear > Controlled > Reusable > Testable
I can work with prompts for ChatGPT, Claude, Gemini, and other generative AI workflows.
This service is suitable for content generation, educational tasks, business workflows, structured outputs, and general AI prompt improvement.
Please send your current prompt, intended task, expected output, and any problems you are experiencing.
Get to know MdHarun
MdHarun
AI Prompt, Workflow and Agentic AI Specialist
- FromBangladesh
- Member sinceSep 2026
Languages
Bengali, English
I design, audit, and improve AI prompts, structured workflows, and agentic AI systems with a focus on clarity, reliability, evaluation, and human review.
My work includes prompt engineering, prompt debugging, LLM evaluation, AI QA, agentic workflow design, educational content, assessment design, and business documentation.
I use a test-and-refine approach to identify failure points, improve instructions, and validate outputs with targeted tests.
My Portfolio
FAQ
What do you need from me to start?
Please send your current prompt, the intended task, the expected output, and any problems you are experiencing. If you have sample inputs or outputs, include them as well.
Which AI tools or models can you review prompts for?
I can review prompts used with ChatGPT, Claude, Gemini, and other generative AI systems. The audit focuses on prompt structure, clarity, constraints, grounding, output requirements, and testability.
Will you rewrite my prompt or only review it?
Depending on the package, I can provide both an audit and an improved version of the prompt. The goal is to preserve what already works and make only the changes needed to improve reliability.
What kinds of problems do you look for?
I check for unclear objectives, missing inputs, conflicting instructions, unsupported assumptions, hallucination risks, weak output controls, formatting issues, duplication, and missing failure-handling rules.
Can you test the improved prompt?
Yes. Standard and Premium packages include targeted test cases designed to check whether the revised prompt handles normal, missing-input, conflicting-instruction, and other relevant failure scenarios.
Can you guarantee perfect AI output or zero hallucinations?
No. AI model behavior can vary by model, context, and input. My service improves prompt structure and testing, but I do not guarantee perfect outputs or complete elimination of hallucinations.
