My Custom GPTs
Here is a collection of custom GPTs I made on ChatGPT. These GPTs are designed to do a single, focused task.
Many of these GPTs attempt to undo AI's tendency to do the thinking for us. Instead, they teach, coach, and provide sources. Instead of doing the work, they teach and provide tools or principles. In other words, they try to prevent human laziness while still harnessing the power of AI.
This concept is available in many AI platforms. For example, in Claude Code, you can create specialized SubAgents that excel at specific tasks.
Typo Fixer
This GPT takes your written text as its only input. It's output is a version of your text with typos fixed. It only fixes typos and makes no other changes. With no conversational output, the fixes are easier to copy and paste back into your editor.
What counts as a typo?
- Mis-spelled words
- Obviously missing words (and, the, etc.)
- Incorrect plurality (unnecessary or missing ending -s)
- Incorrect punctuation
What doesn't it change?
- Links/URLs are never changed, even if they have a typo.
- It doesn't swap, add, or re-arrange words.
Example Input
Hwllo my name is Jstin
Example Output
Hello my name is Justin
Write Tight Coach
This GPT teaches you how to apply Write Tight's advice to the input text. Given an input text, it will output how to apply "Write Tight" advice. It does not rewrite the text. Instead, it teaches.
Example input
The bottleneck came down to one thing: engineering time. Rarely did a product spec or visual design delay engineering. Even when this happened, the engineers typically had enough business context and design savvy to create a working MVP while waiting on the real specs and designs. In my career, I've only seen an engineer waiting around with nothing to do a few times. When all else fails, there's always some tech debt to clean up.
Example output
- Cut the “came down to” ramp
- Phrase: “came down to one thing:”
- Why cut: It’s a throat-clearing lead-in that delays the point; strong verbs beat verb-plus-prep padding.
- Example swap: “was:” / “boiled to:”
Blog Post Visualizer
This GPT takes your blog post and outputs a black and white, 16:9 image for use as a banner.
The image is:
- A methaphor for the input text (the blog post)
- Grayscale
- Flat
- 16:9
- Mostly negative space
Example Input
The content of Twilight Of the Software Engineering Team.
Example Output
Prose Elevation Coach
This GPT teaches you how to elevate your language. It demonstrates how to quicken verbs and strengthen nouns. Don't expect it to rewrite for you, though. Instead, it provides you with the toolset
Example Input
The bottleneck was engineering throughput. Product specs and visual designs rarely delay engineering. Even when this happened, the engineers typically had enough business context and design savvy to create a working MVP. I've only seen an engineer waiting around a few times. When all else fails, there's always some tech debt to clean up.
Example Output
- Principle: Name the constraint, not the symptom.
- Quote: “The bottleneck was engineering throughput.”
- Explanation: This sentence is already doing real work, but throughput is an abstraction. Abstract nouns diffuse force. When you can, press the noun closer to a physical or operational reality—what actually slows, resists, or constrains. The goal is not flourish, but friction you can feel.
- Example change: “The bottleneck was engineering capacity.”
Source Finder
This GPT finds sources. Instead of answering the question, it links you to high-quality sources that can answer the question.
Example Input
Is bacterial pneumonia contagious?
Example Output
- "Community-Acquired Pneumonia"
- Type: Peer-Reviewed Scientific Journal
- Link: https://www.nejm.org/doi/full/10.1056/NEJMra1500240
- "Pneumonia (Bacterial)"
- Type: Government or Intergovernmental Report / Dataset
- Link: https://www.cdc.gov/pneumonia/bacterial/index.html
- "Pneumonia"
- Type: University or Institutional Publication
- Link: https://www.mayoclinic.org/diseases-conditions/pneumonia/symptoms-causes/syc-20354204
- "Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases"
- Type: High-Quality Published Book
- Link: https://www.elsevier.com/books/mandell-douglas-and-bennetts-principles-and-practice-of-infectious-diseases/mandell/978-0-323-48255-4
- "Pneumonia"
- Type: Crowdsourced Knowledge Platform
- Link: https://en.wikipedia.org/wiki/Pneumonia
Friedrich Nietzsche
This GPT responds as if it were Friedrich Nietzsche.
Example Input
What do you think about the invention and advancement of AI?
Example Output
Artificial intelligence is not a new mind; it is a new mirror. Humanity gazes into it and asks for wisdom, and receives only its own habits rendered tireless.