carcall

graphrag-rs. GraphRAG-rs is a high-performance, state-of-the-art Rust implementation of GraphRAG (Graph-based Retrieval Augmented Generation) that builds knowledge graphs from documents and enables natural language querying with configurable entity extraction and local LLM integration

wasm-typst-studio-rs. A WASM-powered Typst Studio built with Rust and Leptos. Features real-time document compilation to SVG/PDF, IEEE template support, dynamic bibliography management (Hayagriva YAML), and an integrated image gallery with IndexedDB storage. Includes syntax highlighting, dual-mode editor (source/visual).

wasm-knowledge-chatbot-rs. 🚀 A blazingly fast, modern chat interface built with Rust/WASM and Leptos, featuring local AI model execution via WebLLM. Privacy-first design with no backend dependencies.

dashboard-studio-rs. Privacy-first, client-side data visualization powered by Rust + WebAssembly. Upload data, create interactive dashboards with 11 chart types, drag-and-drop layout, auto-save, undo/redo—all running in your browser. Zero backend, 100% GDPR-compliant, <500KB bundle.

vecslide. VecSlide is a vector presentation format with synced audio. A .vecslide ZIP contains SVGs and Opus audio via YAML manifest. It compiles to a standalone, zero-dependency HTML file. Built with Rust and Leptos WASM, it shrinks 930MB videos to 13MB files while keeping text selectable and resolution-independent.

lodviz-rs. lodviz-rs is a pure-Rust data visualization library using Leptos and WebAssembly. It excels at automatic Level of Detail (LOD) management, smoothly rendering massive datasets via LTTB and M4 downsampling. With interactive SVG charts, zoom, tooltips, and dynamic themes, it provides a highly performant and flexible solution.

mcp-rustdoc-parser. A Model Context Protocol (MCP) server for parsing and querying Rust documentation. Extracts detailed information about Rust items (functions, structs, enums) from project documentation and serves them via stdio transport. Perfect for integrating Rust documentation into AI-powered development tools.

rust-relations-explorer. A fast, lightweight tool to parse Rust projects into a knowledge graph and run insightful queries over code structure. Generate DOT/SVG visualizations, persist graphs to JSON, and explore code relationships via a clean CLI. ✨

wasm-image-studio-rs. 🖼️ Web-based image editor built with Rust & WebAssembly. Apply filters, adjust colors, and process images in real-time with near-native performance. #Rust #WebAssembly #ImageProcessing

mcp-py-json-doc. A Model Context Protocol (MCP) server that transforms Rust documentation into a queryable knowledge graph, enhancing code understanding and discovery through semantic search and AI integration.

agentic-graphrag-rl-trainer. Reinforcement Learning trainer for LLM function calling using GRPO. Features AST-based reward functions, QLoRA quantization optimized for consumer Graphic Card, and xLAM dataset support. Includes syntax validation, type checking, and semantic correctness evaluation via BFCL AST Score.

wasm-plugin-runtime. A production-grade mini-FaaS/Lambda system built with Rust, WASI 0.2, and the Component Model. Just ~450 LOC delivering hot-reload, Prometheus metrics, Redis/PostgreSQL integration, and Docker deployment. Sandboxed execution with CPU/time limits, <1ms cold starts, 44.5MB optimized image. Clean, secure, extendible.

Rust-MLOps-predictor-RestAPI. The "Prediction" project is an object detection service in Rust and Actix-Web with high performance and robustness.

python-project-analyzer-mcp. A Model Context Protocol (MCP) server that provides tools for analyzing Python projects, extracting library signatures, and retrieving project information. Helps AI assistants understand Python code structure and dependencies.

agent-conductor-rs. A lightweight AI agent orchestrator built in Rust. Features a high-performance Actix-web backend for agent lifecycle management (create, run, pause, cancel) and a real-time WASM dashboard for monitoring/control multi-agent workflows. Production-ready orchestration without external dependencies—fully self-contained, privacy-first, and deployable.

api-llm-dataset-generator. TrainingData-LLM-Generator automates the creation of production-grade synthetic datasets for function-calling LLMs by ingesting OpenAPI schemas and utilizing local inference via Ollama. It employs a self-correcting "Reflexion" architecture to guarantee 100% schema adherence and diverse, context-aware training data without manual scripting.

agentic-devops-assistant. Agentic SRE Assistant: Local-first AI for log analysis, doc RAG, and YAML validation. Built with PydanticAI, DuckDB, and Apache Iceberg.