PromptAudit – AI reply transparency detector

Helps open-source users and contributors spot AI-generated or automated issue replies

Problem

A user noticed GitHub issue replies that read like AI and appeared fully automated, even toward labeled contributors. It's hard to tell whether a human is listening or a bot is brushing you off, undermining trust in open-source communities.

Solution

A browser extension or GitHub App that analyzes language patterns, response latency, and template overlap in issue and PR replies, then shows a probability badge for likely AI automation. It doesn't accuse—it surfaces signals and offers a one-click way to ask for human confirmation.

Users

Open-source maintainers, frequent issue contributors, and corporate open-source compliance teams

MVP

Browser extension that scans issue threads; multi-signal scoring for template-ness, latency, and cross-repo similarity; confidence badge; one-click polite query comment; runs locally without data collection

Revenue

Free browser extension with a team subscription for multi-repo monitoring, dashboards, and exportable reports

Why now

AI auto-replies are spreading fast in open source, but there's no consensus or tooling for transparency; browser extensions and thin APIs let indie developers enter cheaply