Grass Pass: Photograph to Unlock Reddit

Bharath Vankayala — Sun, 11 Oct 2026 20:19:33 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

Grass Pass is a Chromium extension (Chrome, Edge, Brave) that blocks the sites you doomscroll on. To get them back, you have to go outside and prove it

You maintain a block list (by default, X, Reddit, Instagram, TikTok, Facebook, YouTube). Opening one of such websites takes you to a blocked page with a "Go touch grass" button.

A task like "photograph three leaves that look different from each other" is initiated.

You can upload a picture from your phone or use your webcam to take each one.

Every picture is examined by a vision model that runs on your own computer

Demo

The following application is deployed on Vercel
Grass Pass - Demo

Code

Github-Link

How I Built It

The default model is SmolVLM-500M-Instruct, which uses Transformers.js v3 on WebGPU to operate in the browser.

For models like Moondream without a browser build, there is an optional Ollama backend on localhost and an experimental "accurate" tier utilizing Qwen2-VL-2B.

They are all hidden behind a single, tiny VisionBackend interface (init, askYesNo, dispose); thus the remainder of the program is unaware of which one is active.

Small vision models are good at responding to a single yes/no inquiry concerning a single image, but they struggle with counting and reasoning across multiple images. Thus, every image is independently validated in order.

Why Does Open Innovation Matter?

Grass Pass's primary design limitation is privacy, since it requires you to point a camera at your surroundings.

Every picture of your garden, your street would need to be uploaded to someone's server, linked to an account and an API key, and charged for each call if you used a closed vision API.

It's hardly much of a barrier if it stops functioning when the network goes down or your API credits run out.

Open-weight models changed what I could build:

Photos never leave the device. The only network traffic is a one-time download of model weights from Hugging Face.
No server, accounts, or keys. There’s no Grass Pass backend to run or pay for, and nothing to shut down later.
Swappable models. Because the weights are open, the same app can run SmolVLM in the browser, Qwen2-VL on WebGPU, or Moondream through Ollama

The trade-off is that 500M-parameter model is less accurate than frontier model, which is why the design leans on narrow yes/no questions and doesn’t try to defend against a determined cheater.

Personal Perspective

Bharath Vankayala — Sat, 25 Apr 2026 09:19:24 +0000

🌟 My Top Google Cloud NEXT ’26 Highlight

🏆 Chosen Announcement

The Gemini Enterprise Agent Platform
(The complete architectural evolution and rebranding of Vertex AI)

🚀 Quick Summary

Google has effectively "retired" the Vertex AI brand to launch the Gemini Enterprise Agent Platform, a specialized operating system designed not just to host models, but to scale, and orchestrate thousands of autonomous agents. This isn't just a name change; it’s a fundamental pivot from "Generative AI" (producing content) to "Agentic AI" (producing outcomes).

The platform introduces the Agent Runtime, which allows agents to maintain state for days via a new Memory Bank, and the Agent2Agent (A2A) Protocol, an open standard governed by the Linux Foundation that allows Google’s agents to negotiate and execute tasks with agents on AWS, Azure, and Salesforce. Supported by the new 8th Generation TPUs (8t for training, 8i for inference), this platform allows a company to transition from a workforce that uses AI to a workforce that manages a fleet of AI entities.

Why This Stands Above the Rest

From Chatbots to Employees: Most AI updates focus on "smarter answers." This platform focuses on "reliable actions." With Agent Identity, every agent is assigned a cryptographic ID and specific permissions, treating them as first-class corporate citizens rather than ephemeral API calls.
The End of the Walled Garden: The A2A Protocol is a masterstroke. By leading the charge on interoperability, Google is positioning itself as the "connective tissue" of the entire AI economy. An agent in Google Workspace can now natively "hand off" a procurement task to a supply-chain agent running in a competitor’s cloud without human intervention.

🌍 Real-World Impact

Autonomous Logistics: A global retailer can deploy "Vendor Agents" that monitor inventory in BigQuery, negotiate pricing via the A2A protocol with supplier agents, and execute contracts—all within the secure Agent Sandbox.
"Vibe-Coded" Startups: With the integration between Google AI Studio and Cloud Run, founders are now deploying full-stack apps simply by "vibing" (natural language description), letting the Agent Platform handle the underlying microservices, security, and scaling.

🎯 Why It Spoke

We’ve spent the last three years obsessed with "The Model"—the size of the parameters, the length of the context window. But at NEXT ’26, it became clear that the model is now just a commodity "engine." The Gemini Enterprise Agent Platform is the "chassis" and "transmission." It’s the first time I’ve seen a cloud provider provide a serious answer to the question: "How do I actually manage 5,000 agents without them hallucinating my budget away?" It’s the shift from AI as a toy to AI as a utility.

⚡ Final Verdict

The Gemini Enterprise Agent Platform isn't just a product launch; it is the formal declaration that the era of "Individual AI" is over and the era of the "Agentic Enterprise" has begun. Google has built the mission control for the next version of the global economy.

DEV Community: Bharath Vankayala