In this column, I delve into the benefits of prompt engineering tied to a newly launched series of ChatGPT prompt packs, crafted for boosting productivity. OpenAI has recently introduced these packages, aiming to provide pre-designed templates that make it easier for users to integrate generative AI and large language models (LLMs) into their professional tasks.
If you identify as a prompt engineer, it’s essential to familiarize yourself with these innovative prompt packs. They offer convenient templates that can be utilized whenever necessary, allowing you to refine your prompting skills. Additionally, you can support less experienced users who may not be adept at prompt engineering.
Let’s explore this topic further.
This discussion on AI advancements is part of my ongoing Forbes column, where I cover the latest developments in AI, breaking down various complex aspects (for more details, see the link here).
Some of you may remember my previous comprehensive guide on over eighty prompt engineering techniques and strategies (see the link here). Experienced prompt engineers understand that mastering a broad spectrum of well-researched prompting techniques is crucial for maximizing the potential of generative AI and large language models (LLMs).
Prompt Engineering Essentials
An essential aspect of prompt engineering is the precise wording of prompts.
A vital consideration in prompt engineering entails the wording of prompts.
Capable prompt engineers realize that you must word your prompts mindfully to ensure that the LLM gets the drift of what you are asking the AI to do. Sometimes, just an added word or two can radically change what the AI interprets your question or instruction to consist of. Generative AI can be hypersensitive to what you say in your prompts. It is often a touch-and-go proposition.
Plus, there is a potential cost involved. Namely, if you are paying to use an LLM, you’ll be getting an off-target response if your prompt isn’t on-target to your needs, for which you are paying, regardless of whether the LLM grasped your intention or not. As the old saying goes, all sales are final. The same goes for misinterpreted prompts.
Casual users sometimes catch onto this prompt-writing consideration after a considerable amount of muddling around, involving exasperating trial and error. Many users don’t ever become especially proficient in writing prompts. They just enter whatever comes into their minds. That’s probably okay if you are a casual user and only infrequently use AI.
Not so for serious prompt engineers.
Templates As Prompting Suggestions
In a recent posting by OpenAI entitled “ChatGPT For Any Role” on August 7, 2025, various work-related prompting packages or packs were made available, including covering these business-savvy topics:
- Write a professional email.
- Rewrite for clarity.
- Adapt a message for an audience.
- Draft a meeting invite.
- Summarize a long email.
- Create a meeting agenda.
- Summarize meeting notes.
- Create an action items list.
- Prep questions for a meeting.
- Draft follow-up email.
- Identify root causes.
- Compare options.
- Decision criteria.
- Document daily priorities.
- Brainstorm solutions.
- Etc.
Each of those actions provides a templated prompt that you can then fill in with your particular details.
An Example Involving Root Cause Analysis
I’ll go ahead and highlight one of those templated prompts, namely the one entailing root cause analysis.
Suppose that your manager has asked you to figure out why the office keeps getting overwhelmed with work. It seems to be a vexing issue because the volume of staffing is considered on par with the amount of work expected to arise. The manager is despairingly perplexed since the staff ought to sufficiently cope with the prevailing workload.
In a sense, the manager is asking you to undertake a root cause analysis. During a root cause analysis, you first lay out the situation or circumstances that seem to be bogging things down. Next, you attempt to identify key factors that foster problems at hand.
If you’ve never overtly performed a root cause analysis, it takes a bit of elbow grease to do so. Those who often perform root cause analyses might be overconfident and shortchange the process of doing the analysis, including overlooking crucial factors or skimping on the analysis. All in all, having a handy tool to assist in the root analysis can be quite beneficial.
Bringing Generative AI Into The Picture
Aha, a potentially suitable tool would be to leverage generative AI, such as OpenAI’s ChatGPT or GPT-5, Anthropic Claude, Google Gemini, Meta Llama, xAI Grok, and so on.
I mention those various LLMs because the templates that ChatGPT has posted are not especially specific to ChatGPT. You can grab the text of the template and easily use the prompt in some other generative AI. The main aspect of using the templates in ChatGPT is that OpenAI’s postings provide a link next to each template so that you can instantly import the template into ChatGPT (sidenote: a simple cut-and-paste is nearly just as easy).
Here is the root cause analysis template:
- “Analyze the following workplace issue: [describe issue]. The context is that the problem has occurred multiple times. Identify possible root causes and suggest questions to confirm them.”
A user would include a description of the problem at hand. The rest of the text could also be modified as befits the situation.
Focusing On Interactive Collaboration
The beauty of using generative AI is that if the AI ascertains that your description is insufficient to ferret out the root cause, the AI will inform you. You can then proceed into an interactive dialogue about the topic. I mention the notion of carrying on a dialogue or conversation due to the vital realization that LLMs are usually best leveraged in an interactive mode.
Allow me to explain.
A user who assumes that they will merely enter a prompt and get an instantaneous, full-blown answer is not thinking about the use of AI in a proper fashion. The templates get the discussion underway. After that starting point, the user needs to discuss the matter and work with the AI to get to the answer or answers being sought.
A one-and-done mindset needs to be replaced with an interactive, collaborative perspective.
Mentally Exercising Via AI
The odds are that the manager who asked for the root cause analysis is going to ask what the analysis consisted of. They are unlikely to immediately take at face value whatever answer you might convey to them, even if you insist that you have used AI. In fact, assuming you do reveal that you used AI, the manager might get notably suspicious.
We all ought to know these days that AI can make errors. Banner headlines keep pointing out those flaws. In particular, the rise of so-called AI hallucinations entails instances when AI generates falsehoods and posits made-up fiction that doesn’t comport with reality (see my coverage of AI hallucinations at the link here). Anyone using AI, including with these templates, needs to keep their mind engaged and not simply pass along whatever the LLM says.
Getting an accompanying explanation from AI can be a big plus.
When I use templates for prompting, I almost always make sure to add a line that instructs the AI to explain its reasoning or rationale. In the case of this root cause analysis, I would have added a last line that told the AI to do so.
By seeing the explanation, you would have a better understanding of why the AI presumably offers the causes that it has displayed to you. The explanations can become your explanations, though make sure to mull them over and have your own human imprint on them.
Invoking The ChatGPT Use Cases For Work
For those of you who are opting to use ChatGPT for work purposes, and depending upon how you are doing so, the invocation of the templates can take you to a segment known as ChatGPT Use Cases For Work.
The first aspect you’ll see is this type of message:
- “I’m here to help you brainstorm ways to use ChatGPT for Work! I also create custom-tailored prompts for your role. Get started by clicking the button below, then share your job and company.”
Once you’ve clicked on the button, you’ll next see this type of message:
- “Awesome! I’m here to help you brainstorm ways to use ChatGPT in your role at work. To get started, could you please share: Your company name (or the type of organization if you prefer not to name it); Your role/title; A few sentences on what you do, your scope or metrics, and any current projects or challenges. Once you share that, I’ll tailor ideas and prompts specific to your day-to-day work.”
Privacy And Work Confidentiality
I bring up the above so that you’ll be carefully considering whether using public-facing AI is suitable for your work-related purposes.
A lot of people use an LLM such as ChatGPT or Claude without thinking through privacy issues and work-related confidentiality concerns. The employee enters all sorts of data about the company they work for. They might enter data about the customers of the company. This makes sense to the person because they assume that the more details they provide to the AI, the better the AI can answer posed questions.
The problem is that public-facing AI is usually accompanied by an online licensing agreement that says the AI maker can inspect the data that you enter, such as having members of their technical team do so, and they can reuse your entered data when further training the AI (see details via my coverage at the link here). Oopsie, your effort to use the AI for work could give away company secrets, violate privacy laws, and otherwise be a tremendous boondoggle of exposures and legal liabilities.
Firms these days usually publish internal memos telling employees that they are not to use public-facing LLMs and instead use an AI that has been purposefully set up by the company and has numerous privacy provisions. Despite those memos, workers often use public-facing AIs anyway. They do so out of habit, or they never got the memo, or forgot about it, and so on.
Be smart about which AI you use and do not create exposures that you’ll later regret.
In-House Templates For Work And Other
Companies that are at the forefront of using generative AI will often establish their own templates that employees are encouraged to use. The templates resemble the ones that are mentioned above. In addition, they are likely to be more detailed and specific to the company, and they provide background aspects so that you don’t have to repeatedly tell the AI about the nature of the business.
This brings up an alerting point about watching out for potential scams when it comes to free templates for generative AI.
You might remember that when word processing packages first got underway, there were tons of add-ins that you could find on the Internet. People routinely opted to adopt those add-ins. Unfortunately, dastardly evildoers would make available add-ins that could freeze your PC or steal info from your computer. That’s the way the world rolls, sadly.
There are templates online for generative AI that try to do the same evildoing. I mention this so that you will be cautious in searching the Internet to find handy templates related to the use of LLMs. Make sure to carefully inspect the text of the templates. Sneaky twists and turns can be inserted into the template, and you might blindly allow the evildoer to get your use of AI to turn on you.
Prompting For Work Must Be Done Mindfully
Here is a curated set of helpful thoughts about prompting at work:
- Is the generative AI that you are using a public-facing instance or a private version set up for the business?
- Are you mindful of the data that you enter into the generative AI?
- Are there prompting templates available internally that could be useful to you?
- Are there externally available prompting templates that could be useful to you?
- Have you scrutinized the external templates to make sure they are safe for use?
- Do you engage the AI in a dialogue so that you can get a fuller answer and a likely better answer versus a one-and-done approach?
- Do you ask the AI to explain its answers?
- Have you carefully examined whatever the AI says, such that you aren’t going to get jammed up by errors or AI hallucinations?
- Have you reflected on your use of AI to hone your prompt engineering skills?
That last point doesn’t get as much attention as it deserves.
Someone who routinely uses AI should hopefully be improving their prompting skills over time. This typically requires self-reflection on how good or weak your prompts have been. Take a periodic moment of contemplation from time to time to consider and reconsider your prompting style and approach.
As the famed Vince Lombardi remarked, “Practice does not make perfect. Only perfect practice makes perfect.”
