I will deploy a fully local ai agent with tool use and telegram control

Tired of paying per-token API costs? Want an AI agent that never sends your data to the

cloud? I'll deploy a fully autonomous local AI agent on your machine powered by Ollama

or LM Studio that reasons, uses tools, and stays completely under your control.

What you get:

- Autonomous reasoning loop powered by any local model (Llama 3, Mistral, Qwen, etc.)

- Real tool use: file read/write, shell execution, web search, vision analysis

- Telegram integration message your agent from your phone, get live status updates

- Modern web UI for session tracking, live logs, and approving sensitive commands

- Human-in-the-loop safety: destructive commands (rm, git reset) require your approval

- Skill injection system: drop in a SKILL.md to instantly give the agent expert knowledge

(e.g., "Senior DevOps Engineer" or "Data Scientist")

- Swarm mode: spawn sub-agents for research, coding, and writing tasks in parallel

Zero recurring costs. Runs on Mac, Linux, or a home server. Full source code delivered.