</> COMMUNITY FEED
[DISCOVER] [CONNECT] [INSPIRE]
See what the community is building and sharing.
$ Community_Feed.sh
Can Koecher
1d ago

Thien Nguyen
1d ago
Ai Humanizer
This project is an advanced AI-driven text humanization system that leverages Comet for experiment tracking and model optimization, and Weaviate for semantic vector storage and contextual retrieval. The core objective is to transform AI-generated or structured input text into natural, human-like language while maintaining the intended meaning and tone.

Donna Duncan
1d ago
Vibe Coded Code Analyzer
I took the "vibe coding" part of this hackathon very seriously. I gave Cursor this prompt: We are at a hackathon. You are my pair partner. We are going to build an application using Weaviate and Comet. Weaviate Quick Start guide is here: @https://docs.weaviate.io/weaviate/quickstart I have already performed the steps to create the cluster and copied the URL and API key into weaviate_env.txt. Comet's quick start guide is here: @https://www.comet.com/docs/v2/guides/quickstart/ I have completed the signup for Comet and done the pip install comet_ml and comet login steps from the command line Analyze these tools and suggest something we can build together in 90 minutes. We will be judged on how production ready we can make our application, and we will get extra credit for using both toos, but we can use either or both. Any ideas? It suggested "An AI Powered Code Review Assistant". I accepted the suggestion and had it get to work. Failed a bit on the UI, but does appear it used both Comet and Weaviate.

Dillon Devera
1d ago
Ai Ssmart contract auditor
AI smart contract auditor using claude anthropic and opik

Josh Gimenes
1d ago
Prompt-timus Prime
Manually engineering the perfect prompt is a slow, frustrating, and expensive process of trial and error. Prompt-timus Prime automates this by treating prompt optimization as a science experiment. You provide a simple starting prompt and an example of a "good" output. Our system then takes over, creating generations of new prompt variations. It uses Comet to test each variation and score its performance, and a Weaviate vector database to build a "genetic memory" of what works. This creates a powerful feedback loop where the system learns from its own experience, evolving its prompts to become progressively more effective at achieving the user's goal for any given task.

Edinardo Potrich
1d ago

Emtiaz Ahamed Emon
1d ago

Nathan Moos
2d ago
CodeMap
MCP server that generates a dynamic map of a codebase to seed context for agentic editors. Could be leveraged to minimize token/tool-call cost.

2d ago

Dhruvi Kothari
2d ago
