krites
LLM benchmark platform that sends identical prompts to multiple models and compares latency, throughput, cost, and quality side-by-side, with real-time SSE streaming, LLM-as-judge scoring, and interactive analytics charts.
// Open source tools and experiments in Go, Python, and beyond
LLM benchmark platform that sends identical prompts to multiple models and compares latency, throughput, cost, and quality side-by-side, with real-time SSE streaming, LLM-as-judge scoring, and interactive analytics charts.
Minimal AI shell assistant CLI in Go. Type a request in natural language, review the suggested shell command in a TUI, and run it on enter — installable via Homebrew, install script, or go install.
Serverless OpenRouter chat client built with Rust, Leptos, and WebAssembly. The browser connects directly to OpenRouter via OAuth PKCE, with no backend.
Portable, reproducible CPU benchmark for Linux and macOS measuring single- and multi-thread performance across six workloads (primes, sorting, matrix, hashing, JSON), normalized against a reference CPU and exported as JSON.
Automated Discord message cleanup tool that periodically removes messages exceeding a specified age threshold, with bulk deletion support, configurable intervals, and easy deployment via Docker or standalone binary.
Self-hosted LLM observability platform that ingests OpenRouter API call data via webhooks, providing real-time cost monitoring, latency percentiles, and filterable trace exploration backed by DuckDB.
Go implementation of a Model Context Protocol (MCP) server that fetches web pages, extracts readable content, converts HTML to markdown, and supports pagination with robots.txt compliance and proxy support.
Go-based drop-in replacement for npm mcp-remote. Acts as a transparent bidirectional proxy between stdio JSON-RPC and a remote MCP server over HTTP/SSE, with full OAuth 2.1 support in a ~6 MB static binary.
Zero-dependency Go package providing complete bindings for the OpenRouter API with full streaming support, comprehensive error handling, and extensive configuration options for LLM integration.
Interested in collaborating on a project or contributing to open source? I'm always excited to work with fellow developers and explore new ideas.