End the
Duplicate Issue
Triage Nightmare.
An autonomous, cloud-native GitHub Action that leverages semantic embeddings and vector databases to identify duplicate bug reports. Bring your own keys; we handle the automation.
The Stack
Enterprise AI. Microscopic costs.
Uses text-embedding-3-small. Processing a massive 4,000-word issue costs a fraction of a single cent.
Lightning-fast cosine similarity search. The Serverless Free Tier holds up to 2,000,000 issues forever without a credit card.
Containerized execution runs natively in your repository. Zero external servers to manage, deploy, or maintain.
The Data Pipeline
See exactly how the engine processes data from the moment a user hits "Submit new issue."
Smart Extraction
The bot extracts up to 4,000 words from the issue Title and Body. Built-in regex automatically strips out markdown and massive code blocks, ensuring the AI evaluates only the core human intent.
Semantic Translation
The cleaned text is sent securely to OpenAI, which converts the vocabulary into a 1,536-dimensional coordinate vector. This neutralizes differences in wording (e.g., "log in" vs. "authentication").
Search & Flag
Pinecone executes a Cosine Similarity search against your repository's memory index. If a past issue matches with > 65% similarity, the bot posts a friendly comment flagging it for human review.
Ready to automate your repo?
Copy this workflow into .github/workflows/duplicate-checker.yml to activate the AI.
name: Issue Duplicate Checker
on:
issues:
types: [opened, edited]
jobs:
check-duplicates:
runs-on: ubuntu-latest
permissions:
issues: write
steps:
- name: Checkout Repository
uses: actions/checkout@v4
- name: Run AI Duplicate Detector
uses: Shashank-8p/issue-detector@main
with:
openai_api_key: ${{ secrets.OPENAI_API_KEY }}
pinecone_api_key: ${{ secrets.PINECONE_API_KEY }}
pinecone_index_name: ${{ secrets.PINECONE_INDEX_NAME }}
github_token: ${{ secrets.GITHUB_TOKEN }}