Promptaus: An AI-Powered Prompt Management Tool
This article discusses Promptaus, the first product created using AI. The creator emphasizes its serious intent to solve a specific problem related to prompt management and utilization, not just a superficial experiment.
The Problem: Managing and Utilizing Prompts Efficiently
The Need for Prompt Management
The article highlights the increasing importance of prompts in interacting with AI. Prompts are considered the closest AI usage technique to the user, making their audience potentially vast.
Inefficiencies of Traditional Storage
The speaker identifies a common problem: storing and retrieving prompts. Manually storing prompts in notebooks and copying/pasting them is inefficient.
The Solution: AI-Assisted Prompt Selection
The core idea is to enable AI to automatically select and call prompts based on user needs, streamlining the process.
Promptaus: A Web-Based Solution
Functionality
Promptaus is a web application designed to store, manage, and organize prompts. Users can save prompts, assign titles and labels, and manage them in detail.
MCP Protocol Integration
Promptaus leverages the MCP protocol to facilitate AI-driven prompt selection. This involves:
-
The client model retrieving a list of available prompts (titles and labels) from the server.
-
The model analyzing the list to determine the best prompt for the given situation.
-
The model requesting the complete content of the selected prompt from the server.
-
The server sending the prompt content back to the client.
Simple Process
The entire process is described as simple: get the list, select the prompt, and get the complete content.
Development Process and Tools
Rapid Development
The development time for Promptaus was surprisingly short, only about five or six days from start to launch.
UI Design
The initial UI was generated using V0 but was later redesigned using Keynote, highlighting the creator's comfort with using familiar tools.
Key Technologies
The following technologies were crucial in Promptaus's rapid development:
-
Cursor: Used with cloud models to automate code completion.
-
Vercel and SuperBase: Provided a suitable platform for the early stages of development, with SuperBase being particularly praised for its convenience.
-
OpenAI Models: Cloud 3.5 + Gemini Pro was used in the beginning. OpenAI O3 was used to address bugs. Finally, Cloud 4 was used with cursor.
Core Modules
The project consists of four core modules, including a Google account-based user authentication system and AI-powered "smearing" functionality using the OpenAI model for prompt management.
Overcoming Challenges
Minimal Coding
The creator emphasizes that they wrote very little code themselves; most of the development was done by Cursor.
Model Selection
Initially, Cloud 3.7 and Gemini 2.5 were used. OpenAI O3 was used when there were bugs. The emergence of Cloud 4 with a discount in Cursor significantly accelerated the development process.
Recommendation
Cloud 4 is highly recommended as the most powerful programming model currently available.
Cost-Effectiveness
The total cost for developing Promptaus was around $100 over approximately one week.
Future Plans and the Changing Landscape
Future Products
The creator plans to develop more products using a similar AI-assisted approach, aiming to launch three to six products before the end of the year.
Shifting Distribution Mode
The speaker believes that the distribution mode of traffic is undergoing a dramatic shift with the move from PageRank to AgentRank. Websites will need to be remade.
Importance of Connecting with AI
It's crucial to connect with AI using MCP or future protocols to capture traffic, as traffic will first flow to those who are prepared.
Lowered Product Creation Threshold
The threshold for creating products has significantly dropped. While knowing how to code isn't necessary, programming, structural, product, and commercial thinking are still vital.
The Rise of the Non-Coding Creator
A group of people with strong product sense and problem-solving skills who were previously blocked by coding barriers can now leverage AI tools to create and innovate.
Focus Shift
The speaker has reduced content creation to two days a week to dedicate the remaining time to coding.
Community and Sharing
Community Engagement
The entire development process, from initial thoughts to challenges and results, is being shared within a community.
Value Proposition
The community provides value by sharing how the creator thinks, acts, and even the financial outcomes of their projects.
Newtype Community
Individuals interested in AI, becoming super individuals, and finding the right connections are encouraged to join the Newtype community.