Personal Telegram agent platform
AverageClaw
Your own agent. Your own container.
Set up Telegram agents for individual users and run them separately. The dashboard shows each model and runtime state with start, stop and restart controls.
12Commits
Built, one commit at a time
As of About this number
Counts the commit history of the associated repositories on their default branches. The number includes all contributors up to the date shown.
01 / What it does
AverageClaw connects a personal Telegram bot to its own agent instance. Each instance gets a separate Docker container rather than sharing one running agent across all users. I built the platform from the setup flow through to container management.
02 / What I built
Separate agent instances
Implemented container management with separate instances and configurations for individual users.
Guided setup
Built a multi-step setup flow for the Telegram bot, user association and model selection.
Visible runtime state
Added dashboard controls to start, stop and restart agents, display their models and access chat, tools and memory.
Accounts and ownership
Connected server-side user accounts, password hashes and signed sessions to a relational database.
Screenshots
