// about hublens

HubLens is a daily digest of trending open-source repositories, automatically discovered from GitHub and Hacker News. Each day, our pipeline surfaces the most relevant projects gaining momentum — scored, ranked, and summarized so you can stay current without the noise.

We built HubLens because tracking OSS trends manually is time-consuming. Instead of scrolling through countless repos, you get a curated snapshot: what's rising today, why it matters, and how you might use it.

Data is collected daily from two sources:

• 01GitHub Trending — repositories with significant star growth over the past 24 hours, fetched via the GitHub API.
• 02Hacker News — Show HN and notable OSS-related posts from the HN feed.

Raw data is processed by Vertex AI (Gemini) on Google Cloud to generate English and Chinese summaries, extract use cases, and assign category tags. No human editorial judgment is applied — all content is AI-generated from the repository metadata and README.

In addition to summaries, the AI extracts two types of metadata:
- Category: A broad classification (e.g., AI, DevOps, Frontend) chosen from a fixed list to describe the project's primary domain.
- Tags: Dynamic keywords (e.g., #OpenAI, #Docker, #CLI) extracted from the repository's features.

Each repository on HubLens displays a source badge indicating how it was discovered. Here's what each badge means:

Beyond the daily trending digest, HubLens offers tools for deeper exploration:

• 01Pin & Compare — pin up to 5 projects and compare them side-by-side across stars, score, category, tags, summary, and more.
• 02Star Change Tracking — see how many stars each project gained or lost since yesterday, displayed alongside the current star count.
• 03Topic Hubs — browse projects by technology topic (e.g. AI, React, Rust). Only topics with 10+ articles in the last 90 days are listed.
• 04Keyboard Navigation — vim-style shortcuts (j/k/gg/G/o/p/c/?) for fast browsing on desktop.
• 05Public API — access HubLens data programmatically via our REST API. Free, no authentication required.
• 06MCP Server — use HubLens data from AI tools like Claude Code. See the API page for setup instructions.

Each repository is assigned a HubLens Score — an objective, algorithm-derived number that reflects recent momentum, not absolute popularity. The score is calculated across four equally weighted dimensions (25% each):

• 01Popularity — Relative ranking based on current stars against other trending projects in a 30-day window.
• 02Momentum — Relative ranking of the trending score (growth rate).
• 03Usefulness — Qualitative assessment by AI on a 5-point scale (20-100).
• 04Innovation — Qualitative assessment by AI on a 5-point scale (20-100).

To ensure emerging projects are not buried by massive repositories with long-accumulated stars, we apply a New Project Boost. Projects created within the last 14 days receive a +15 point bonus, and those within 60 days receive a +10 point bonus.

The breakdown on each repo page shows these dimensions on a 5-point visual scale (very weak / weak / moderate / strong / very strong).

The score is a relative ranking tool to help you discover what's gaining traction right now. It does not represent the quality, maturity, or long-term viability of any project, and should not be interpreted as an endorsement or ranking of OSS value. Every project has its own merits — we encourage you to explore each one on its own terms.

HubLens aggregates publicly available data from third-party platforms (GitHub, Hacker News). We do not guarantee the accuracy, completeness, or timeliness of the information displayed.

AI-generated summaries may contain errors or omissions. Always refer to the original repository for authoritative information. HubLens is not affiliated with GitHub, Hacker News, or any of the projects listed.

Use of this site is at your own risk. We accept no liability for decisions made based on information presented here.

HubLens participates in Google AdSense. Ads displayed on this site are served by Google and comply with Google Publisher Policies.

Advertisements are clearly distinguished from editorial content. We do not endorse products or services shown in ads. For ad-related inquiries or to opt out of interest-based advertising, use the AdChoices link in the ad unit or visit Google Ad Settings.

Have a question, found an issue, or want to report incorrect content? Reach us by email:

[email protected]