Vibe Coding Discoverai-microcore
View on GitHubA handy lib for smooth interaction with large language models (LLMs) and crafting AI apps.
MicroCore is a minimalist Python library of LLM and vector-DB adapters that makes providers switchable via config while keeping app code unchanged. It includes prompt templating, streaming, embeddings search (Chroma/Qdrant) and LLM-agnostic MCP tool integration.
Use Cases
Call any LLM with one unified Python APISwap between OpenAI, Anthropic, Gemini, Azure and local models via configConnect MCP tools to models that lack native MCP supportSemantic search over documents with ChromaDB or QdrantRender reusable Jinja2 prompt templatesBuild Telegram bots backed by LLMsBuild streaming web chat UIs with FlaskRun inference through CLI tools like claude or geminiAdd memory and RAG to chat applicationsGenerate images from promptsTrack token usage and metrics per requestPrototype AI apps with minimal boilerplate
Built With
- Language
- Python
- Frameworks
- Jinja2 · FastMCP · MCP · OpenAI SDK · Pydantic · python-dotenv · tiktoken · ChromaDB · Qdrant · Hugging Face Transformers · PyTorch · aiogram · pyTelegramBotAPI · Flask · Azure Identity
Tags
llm · python · mcp · rag · vector-database · prompt-templates · provider-agnostic · embeddings · semantic-search · streaming · jinja2 · llm-adapters · chat-completion · tool-calling · minimalist