Smart Technology Detection

When AI generates a post from your GitHub activity, it needs to know which technologies you used. Guessing leads to inaccurate tags. Our new rule-based detection system solves this.

The Problem

Previous tag generation relied entirely on the AI model's interpretation. This led to:

• Tags for technologies only mentioned in passing
• Missing tags for tools that were central to the work
• Inconsistent tagging across different AI providers

The Solution

We built a rule engine with 267 detection rules that analyze your commits, pull requests, and code reviews for concrete evidence of technology usage.

How Detection Works

Each rule maps patterns to a technology:

• File extensions — A .rs file means Rust, .swift means Swift
• Configuration files — package.json signals Node.js, composer.json signals PHP
• Import statements — use Illuminate\ indicates Laravel
• Framework patterns — Migration files, route definitions, component structures

The rules only tag technologies with direct evidence. No guessing, no hallucination.

Explicit Language Selection

You can now set the primary programming language for each post directly in the generator. This appears as a dedicated field, separate from tags, giving you precise control over how your content is categorized.

Tighter Tag Generation

Tags now follow strict rules:

• Only include technologies with direct evidence in the source material
• Remove vague or generic tags
• Limit to the most relevant technologies for each post

The result is cleaner, more accurate metadata that improves discoverability and credibility.

Always Improving

The rule set grows as we encounter new frameworks and tools. Each rule is tested to minimize false positives while catching real usage patterns.

Better detection means better content. Your posts now accurately reflect the technologies you work with.