This product was not featured by Product Hunt yet. It will not yet shown by default on their landing page.
Product upvotes vs the next 3
Waiting for data. Loading
Product comments vs the next 3
Waiting for data. Loading
Product upvote speed vs the next 3
Waiting for data. Loading
Product upvotes and comments
Waiting for data. Loading
Product vs the next 3
Loading
Storypointinator
AI sprint estimation that reads your tech debt.
Sprint planning shouldn't be a guessing game. Storypointinator is an open-source VS Code extension that reads your active codebase, detects hidden tech debt, asks clarifying questions, and generates mathematically consistent Fibonacci story points—right in your IDE.
Hey Product Hunt! 👋 I’m Brooks, the creator of Storypointinator.
If you’ve ever worked in software, you know the absolute dread of the "3-point ticket trap."
A Product Manager brings a vague 2-sentence feature request to sprint planning. The developers guess it will take a few days, so they throw up a "3". Halfway through the sprint, a developer opens the codebase and realizes the legacy auth.ts file is a 2,000-line undocumented nightmare.
The sprint blows up. Budgets get stretched. Everyone is frustrated.
Estimating a software ticket without looking at the codebase is like quoting a kitchen remodel without looking behind the drywall. So, I built Storypointinator—an open-source AI Technical Product Manager that lives directly inside your VS Code sidebar.
Here is how it fixes sprint planning:
🕵️♂️ Reads the Tech Debt Blast Radius: It uses VS Code APIs to read your actively open tabs and directory structure. If it sees you are touching a horrific legacy file, it mathematically increases the story point estimate to account for the tech debt.
💬 Agentic Q&A Loop: It doesn’t just blindly guess numbers. If your feature request is vague (e.g., "Make the table faster"), the AI will ask you clarifying questions until the requirements are rock solid.
🔌 MCP Integration: It supports the new Model Context Protocol (MCP). It can connect to your local Design System to see if a UI component already exists before estimating frontend complexity.
📝 Jira-Ready Output: Once the scope is clear, it generates a perfectly formatted Markdown ticket with BDD Acceptance Criteria, out-of-scope assumptions, and a Fibonacci estimate.
🔒 Purely Local Context
Security is huge. You don't have to zip up your proprietary enterprise repo and send it to a random cloud server. Your code stays local, and the extension only fetches the specific context it needs to run.
I’d love for you to download it, throw your vaguest feature requests and messiest code files at it, and let me know what number it spits out! Happy to answer any questions about the LangGraph architecture, our custom MCP integration, or how we built the React/Vite UI in a webview!
What is the worst estimation disaster your team has ever had? Let me know below! 👇
