Generative AI can make processes in university administration simpler, faster and sometimes even a little more creative. For example, it supports the formulation, structuring, translation or analysis of content - often in just a few seconds.
On this page you will find examples that you can try out directly. No prior knowledge required - simply copy the prompt, paste it into an AI tool and see what happens.
Important: Pay attention to data protection and information security. Avoid using sensitive and personal data. When using AI, you are responsible for the content you enter and for the use of the results generated. Follow the KIT guidelines.
The prompts used here are taken from the course "AI skills in university administration" from the KI-Campus. For a more in-depth discussion of the topic, you can view the course free of charge on the official website of the KI-Campus.
Long or complex e-mails from students can be answered quickly, correctly and in a friendly manner with this prompt. The fixed structure ensures clarity: first a summary (approx. 80 words), then the actual reply (approx. 150 words) and finally a short subject (approx. 10 words). If information is missing, this is addressed politely. This creates consistent, comprehensible and service-oriented responses - ideal for student services.
The prompt is suitable for use in e-mail processing, formulating reply templates or quality assurance of texts.
# Scenario:
A student inquiry should be answered quickly, correctly and empathetically.
# Role & tone:
You are an employee of a university (e.g. in student services). You answer in a service-oriented, friendly manner and in clear language
# Task & goal:
Answer a comprehensive request completely and clearly. If information is missing or unclear, point this out in a friendly manner.
# Structure & scope:
- Summary (approx. 80 words)
- Answer (approx. 150 words)
- Subject (approx. 10 words)
# Format specification:
Summary:
Answer:
Subject:
# Incoming request:
<<< [Query text and insert] >>>
# Technical key points / content for answering:
<<< [Name key points and content for answering that the AI bot should use] >>>
With this prompt, you can quickly create clear, appealing and addressee-oriented information letters or invitations. It defines your role as a communications employee:in a university, specifies the typical structure of official letters (salutation, introduction, details, conclusion, greeting) and takes into account key contextual factors such as occasion, target group, location and date. This ensures that the style and form of address remain consistent, professional and friendly.
Areas of application range from event announcements and invitations to information emails about deadlines or changes in the study program to cover letters for official documents.
# Scenario:
A formal letter (e.g. information letter or invitation) should be formulated clearly and attractively.
# Role & tone:
You are a communications employee:in a university and formulate in a friendly, professional and addressee-oriented manner.
# Context: Occasion Target group Location Date
# Task & goal: Create a complete information letter or invitation in continuous text.
# Structure:
- Salutation
- Introduction
- Detailed information
- Conclusion
- Greeting
# Length: 150-200 words
# Format: Continuous text, "you" address
# Output format:
- Subject:
- Email:
# Input source: <<< [insert details of occasion, target group, location, date] >>>
With this prompt, extensive documents - such as legal texts, minutes or statements - can be quickly reduced to their essential content. The AI takes over the first level of analysis: it filters out the most important points, summarizes them in a structured way and creates a practical task list for the administration. In this way, complex content is transformed into a clear basis for action.
# Scenario:
Extract key statements and to-dos from a multi-page PDF document.
# Role & tone:
You are a legal expert in the university administration.
You formulate factually, precisely and with a view to administrative implementation.
# Context:
Only use excerpts or content of the document. Actively request document if necessary.
# Task & goal:
- Identify key statements
- Create task list
# Structure:
1. Summary (max. 10 points)
2. Task table
3. Warning (optional)
# Length:
Max. 250 words
# Format specification:
- Summary: - ...
- Table: | Task | Role | Deadline |
- Warning: <1 sentence
# Input source:
<<< [insert text excerpt] >>>
With this prompt, German-language sections from official documents - such as examination regulations - can be translated into English with legal certainty. The role is clearly defined: The AI acts as a specialist legal translator and delivers precise, stylistically appropriate translations without any loss of information. Technical terms are translated exactly, and a short footnote explanation is provided for untranslatable terms (maximum 20 words). The fixed structure ensures consistent, legally comprehensible results - ideal for administrative staff working in an international context.
# Scenario:
A paragraph of an examination regulation is to be translated into English with legal certainty.
# Role & tone:
You are a specialist legal translator (German to English).
You translate precisely, with legal certainty and without loss of information.
# Context:
Only the paragraph provided is to be translated - no abbreviations, no additions.
# Task & goal:
Translate terms exactly.
Add footnotes to untranslatable terms (max. 20 words).
# Structure:
- Translation
- Footnotes (if required)
# Format:
- Translation:
- Footnote:
- Explanation:
# Input source:
<<< [German paragraph] >>>
This prompt creates a professional, gender-appropriate and reader-friendly job advertisement for university positions in just a few minutes. It is specially tailored to the role of recruiters and guides you step by step through the most important building blocks: title, tasks, profile, benefits and closing sentence. The language remains clear, inclusive and respectful when addressing you. The result can be used directly for websites or job portals.
# Scenario:
Formulate a quick, inclusive job advertisement for the position of a consultant (E 13 TV-L).
# Role & tone: You are a personnel officer at a university. Your wording is clear, structured, gender-appropriate and uses respectful language.
# Context: All relevant information about the position (e.g. tasks, requirements, general conditions).
# Task & goal: Create an appealing job advertisement for a Referent (m/f/d), E 13 TV-L.
# Structure:
- Title
- Tasks
- Profile
- We offer ...
- Closing sentence
# Length: Maximum 320 words
# Format specification: Continuous text, you address, formulated in a gender-appropriate way
# Output format: Job advertisement:
# Input source:
<<< [insert job details] >>>
With this prompt, you can transform large budget tables into comprehensible analyses that are easy to understand even without in-depth knowledge of figures. The AI automatically identifies the three largest budget variances (actual vs. plan) over 10%, explains the background in layman's terms and provides specific recommendations for further action.
Ideal for controlling teams, budget discussions or strategic meetings: The results appear in a structured deviation table, supplemented by practical recommendations and a compact management summary. This gives managers and colleagues without a focus on figures exactly the information they need to make quick and well-founded decisions.
# Scenario:
The three largest budget variances over 10% are to be identified and explained in layman's terms.
# Role & tone: You are a controlling analyst and formulate the information in a fact-based, understandable and target group-oriented way.
# Context: The basis is a budget CSV file with a plan/actual comparison.
# Task & goal:
1. Identify and explain the top 3 variances
2. Make concrete recommendations
3. Create summary
# Structure:
1. Deviation table
2. Recommendations
3. Summary
# Length:
Summary: max. 200 words
# Format specification:
- Deviations: | Cost center | Diff % | Cause |
- Recommendations: - ...
- Summary:
# Input source:
<<< [insert CSV excerpt or numerical values] >>>
This prompt supports procurement managers in evaluating laptops or other devices objectively, comprehensibly and in line with the budget. Based on your specifications - such as budget limits and sustainability criteria - the AI creates a structured catalog of criteria.
The output is in a clear table format with minimum requirements, measurement methods and justifications. Optional additional services such as support or delivery conditions are also taken into account. This allows you to create transparency and comparability, which are particularly important for public tenders and budgetary requirements.
# Scenario:
25 laptops (≤ €1,200 per device) are to be objectively evaluated.
# Role & tone of voice:
You are a procurement expert in the IT sector and formulate precise, technically sound and comprehensible arguments.
# Context:
Consider both budget requirements and sustainability aspects.
# Task & goal:
Create a list of criteria for a structured needs assessment.
# Structure:
1. Minimum specifications as a table
2. Justifications
3. Additional services (e.g. support, warranty, delivery)
# Length:
Maximum 300 words
# Format specification:
- Table: | Criterion | Minimum value | Measurement method |
- Justifications: - ...
- Additional services: - ...
# Input source:
<<< [insert requirement or technical framework data] >>>
This prompt creates a clearly structured agenda for complex meetings - for example on personnel or IT budget issues - in just a few minutes. The AI creates a clear 90-minute agenda based on your documents (e.g. minutes, budget plans, topic requests) and adds a list of preparation tasks for all participants.
The output is in a practical table with time, topic, objective and method - perfect for committee meetings, steering groups or project meetings. Thanks to the clear structure, nothing is forgotten, planning becomes more transparent and the meeting more efficient. Even without prior technical knowledge, this prompt can be used immediately and saves valuable time in day-to-day administration.
# Scenario:
A 90-minute meeting (topic: personnel & IT budget) is to be planned in a structured manner.
# Role & tone:
You are the speaker, responsible for organizing the meeting and taking minutes.
You plan efficiently, in a structured and targeted manner.
# Context & sources:
- Previous minutes
- Budget plan
- Participants' topic requests
# Task & goal:
Create a well thought-out agenda and a list of preparatory tasks.
# Structure:
1. Agenda table
2. Preparatory tasks
# Length:
Maximum 250 words
# Format specification:
- Table: | Time | Topic | Objective | Method |
- Tasks: - ...
# Input source:
<<< [insert meeting information, previous minutes, budget plan, topic requests, etc.]. >>>
This prompt can be used to create AI-supported decision templates for university management. Different data sources - such as evaluation reports, surveys or budget plans - are automatically evaluated and condensed into a structured basis for decision-making: a brief overview of facts, possible courses of action and a well-founded recommendation in a management-oriented style.
Prompt is suitable for a wide range of decision-making processes in university administration, e.g:
- Program development: evaluation of evaluations and resource overviews
- Organizational development: Deriving structural changes from feedback and objectives
- Investment planning: Decision support for infrastructure purchases and funding
Note: This AI-supported template only serves as a decision-making aid. Check the results carefully, make decisions independently and document the use of AI in accordance with the KIT guidelines.
# Scenario:
Several sources of information (evaluation report, survey, budget plan) are to be condensed into a decision template.
# Role & tone: You are an analyst and formulate clearly, fact-based and management-oriented.
# Context:
- Evaluation report
- Survey results
- Budget plan
# Task & goal:
Derive a well-founded recommendation for a decision.
# Structure:
Initial situation → Facts → Options (table) → Recommendation
# Length:
300-400 words
# Format specification:
- Present options as a table
- Rest as structured continuous text
# Output format:
Decision template:
# Input source:
<<< [Insert excerpts from report, survey and budget] >>>
This prompt transforms the AI into a constructive sparring partner: it takes on the role of a prompt engineer and coach that analyzes existing prompts, makes suggestions for improvement and creates an optimized variant. The three-part structure - suggestions, new prompt, justification - not only helps with the specific task (here: "Summarize examination regulations"), but also sharpens your own prompting skills in the long term.
# Scenario:
Ask the AI for feedback on an existing prompt and for an optimized variant.
# Role & tone: You are a prompt engineer and coach. You give well-founded, constructive feedback in clear language.
# Context: Original prompt: "Summarize examination regulations"
# Task & goal:
1. Name three concrete suggestions for improvement
2. Suggest an optimized prompt
3.Briefly justify the improvement
# Structure:
- Improvement 1-3
- Optimized prompt
- Justification
# Format specification:
- Bullet points allowed
- Before/after clearly distinguishable
- Optimized prompt as markdown block
# Output format:
- Improvement 1-3: - ...
- Optimized prompt: markdown ...
- Justification:
# Input source:
<<< [Insert original prompt here] >>>
With this prompt, it is possible to create an effective and customized prompt step by step, even without previous experience. The AI takes on the role of a prompt engineer and guides you through a clearly structured co-creation process: first there is an initial draft with ideas for improvement, then open questions are clarified and the prompt is optimized together.
The process is particularly suitable for employees in university administration who want to develop specific texts or solutions with generative AI (e.g. ChatGPT) - from email formulations to concept drafts. The dialog-based process produces high-quality prompts in a short space of time that can be used immediately.
# Your role:
I would like you to become my prompt engineer. Your goal is to help me create the best possible prompt for my needs through constant repetition and improvement. The prompt will be used by you, the AI system, at the end.
# Work steps:
**Begin:**
Your first response begins with a greeting. You point out that you are a prompt engineer and ask me what the prompt should be about.
**Step 1:**
You create three sections:
- *Section A - Draft prompt:* An understandable, concise prompt in the form of a request to an AI system.
- *Section B - Suggestions for improvement:* Ideas on what could be added or sharpened in the prompt.
- *Section C - Queries:* Specific questions for me to improve the prompt.
**Step 2:** I answer the queries. You use my information for a revised version and start the next cycle with step 1.
# Goal: A fully optimized prompt that can be executed directly by the AI system.
# Input source:
<<< [wait for start signal from me] >>>
With this prompt, you can turn generative AI into a personal quiz tutor that responds flexibly to the subject area, level of difficulty, language and level of knowledge. Multiple-choice tasks with individual feedback create a motivating learning format that builds up and consolidates knowledge in a targeted manner.
Ideal for staff development in university administration (e.g. data protection, IT security, administrative law), for departmental training (e.g. third-party funding management, examination law) or for university didactic self-study courses. The Quiz Tutor can also be used to evaluate course content after training courses.
The structured process ensures that learners actively participate, use mistakes as learning opportunities and immediately receive appropriate explanations - for lasting learning success.
# Your role:
You are a quiz tutor who sets me multiple choice tasks. The tasks serve as training tasks for me in terms of formative assessment. This means I can take quizzes from you on any subject area, e.g. teaching or administration. It is very important to you that I learn something new by completing the tasks and receiving your feedback.
# Your work task:
You formulate multiple choice tasks in a language that I specify to you as part of the work steps described below. You support me if I have answered a task incorrectly.
# Format of the multiple choice tasks:
Each task begins with a question stem that ends with an indication of how many answers are correct.
Example structure:
**Question stem:** [here is the wording of the question stem]
**Option 1:** ...
**Option 2:** ...
**Option 3:** ...
**Option 4:** ...
**Option 5:** ...
*Important:* Do not mark any correct answer options.
# Steps:
1. You ask me about the topic of the tasks.
2You ask me about the level of the tasks.
3. You ask me about the language in which the tasks should be set.
4. You ask me for a knowledge base.
5. You set a task.
6. I answer. You give feedback (correct/incorrect + explanation).
7. Then repeat step 5.
# Restriction: If I give you a knowledge base, please do not ask any questions about the formal structure (e.g. no questions about chapter numbers or similar).
# Input source:
<<< [Wait for my start signal, then start step 1] >>>