MEROBOT - PERSONAL AI ASSISTANT
Feb 2026 - Mar 2026 • CLI / API
SNAPSHOT
An opinionated personal AI assistant built from scratch with Python, featuring iterative tool-calling, message-bus architecture, and sandboxed execution - enabling complex multi-step tasks through LLM-driven automation.
THE PROBLEM
Building a robust personal AI assistant requires integrating multiple components: LLM providers, communication channels, tool execution, and maintaining conversation context, all while ensuring security, extensibility, and performance.
THE RESULT
Developed a fully functional AI assistant with 8 built-in tools, supporting multiple LLM providers and Telegram integration, demonstrating expertise in async Python development, API design, and agent-based architectures.
HOW IT WORKS
MeroBot is a comprehensive AI assistant that bridges communication channels (currently Telegram) with large language models through an async-first message-bus architecture. The system implements an iterative agent loop where the LLM can chain up to 10 tool calls per message, enabling sophisticated multi-step reasoning and task execution. Key technical achievements include: - **Modular Architecture**: Clean separation of concerns with dedicated layers for agents, communication handlers, LLM providers, and tools - **Tool Ecosystem**: 8 built-in tools covering file operations, web search/scraping, code execution, database queries, and sub-agent delegation - **Provider Flexibility**: Pluggable LLM backend supporting OpenAI-compatible APIs, LiteLLM (100+ providers), and direct HTTP implementations - **Security & Sandboxing**: All file and code operations confined to a configurable workspace directory - **Async Design**: Full asyncio implementation with message queues for scalable, non-blocking operations - **Extensibility**: Well-documented ABCs and interfaces for adding new channels, providers, and tools