Repolode
Submit
emanueleielo

emanueleielo/ciana-parrot

Self-hosted AI personal assistant that chats via Telegram, runs scheduled tasks, and connects to MCP servers and host tools.

View on GitHubWebsite
Stars
77
Forks
19
Open issues
1
Last commit
17 days ago

What is it?

CianaParrot is a self-hosted AI personal assistant built on DeepAgents, LangChain and LangGraph. The agent runs inside a Docker sandbox and reaches the host operating system through a gateway of secure bridges. Users interact with it mainly through Telegram, and the agent can also create cron, interval and one-shot tasks from chat.

The project supports multiple LLM providers, including Anthropic, OpenAI, Google Gemini, Groq, Ollama, OpenRouter and vLLM. Several model tiers can be configured so the agent switches models during a conversation. Host bridges cover tools such as Spotify, Reminders, Obsidian and 1Password, and external MCP servers can be connected. The whole setup is configured through a single YAML file.

Skills are added by dropping a folder into the skills directory, where they register automatically as agent tools. Memory is kept in Markdown files that the agent can update itself. Deployment is Docker-based, with make targets for build, run and gateway startup, and an install script handles prerequisites and API key setup.

Key features

  • 01Chat with the agent through Telegram, with a pluggable channel architecture for other channels
  • 02Supports Anthropic, OpenAI, Gemini, Groq, Ollama, OpenRouter and vLLM as LLM providers
  • 03Cron, interval and one-shot scheduled tasks that can be created via chat
  • 04Host gateway bridges to tools such as Spotify, Reminders, Obsidian and 1Password
  • 05Skills auto-register as agent tools when placed in the skills folder
  • 06MCP server integration and Markdown-based persistent memory

Use cases

  • Running a private AI assistant that you message from Telegram on your own server
  • Scheduling recurring reminders or automated checks that the agent runs on its own
  • Connecting a local model through Ollama or vLLM while keeping data on your infrastructure
  • Extending an assistant with custom skills or external MCP tools

Quick start

  1. 1

    Clone the repository

    git clone https://github.com/emanueleielo/ciana-parrot.git
  2. 2

    Enter the project folder

    cd ciana-parrot
  3. 3

    Copy the env file and add your API keys

    cp .env.example .env
  4. 4

    Build and start the Docker containers

    make build && make up

Things to know

  • Requires Docker and an LLM provider API key, or a local model setup.
  • Telegram is the only channel available out of the box.
  • Host gateway features depend on specific macOS and CLI tools.

Topics

#ai-agent#personal-assistant#telegram-bot#self-hosted#mcp#langgraph#docker#scheduled-tasks#ai-assistant#automation

FAQ

What is ciana-parrot?
Self-hosted AI personal assistant that chats via Telegram, runs scheduled tasks, and connects to MCP servers and host tools. CianaParrot is a self-hosted AI personal assistant built on DeepAgents, LangChain and LangGraph. The agent runs inside a Docker sandbox and reaches the host operating system through a gateway of secure bridges. Users interact with it mainly through Telegram, and the agent can also create cron, interval and one-shot tasks from chat.
Is ciana-parrot free to use? What license does it have?
ciana-parrot is open source under the MIT license. Check the license terms before using it in commercial products.
Is ciana-parrot still maintained?
Actively maintained. The last commit to ciana-parrot was on Oct 5, 2026.
What can I use ciana-parrot for?
Running a private AI assistant that you message from Telegram on your own server Scheduling recurring reminders or automated checks that the agent runs on its own Connecting a local model through Ollama or vLLM while keeping data on your infrastructure Extending an assistant with custom skills or external MCP tools
How do I install ciana-parrot?
Clone the repository: git clone https://github.com/emanueleielo/ciana-parrot.git Enter the project folder: cd ciana-parrot Copy the env file and add your API keys: cp .env.example .env Build and start the Docker containers: make build && make up
Where can I get ciana-parrot?
The source code and installation instructions are on GitHub: https://github.com/emanueleielo/ciana-parrot

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects