I will lock down your self hosted ai agent on your vps

Your AI agent holds your API keys and often your email and a shell. If it is reachable from the internet, anyone who finds it can use them.

I'm Greg, a Linux admin with 26 years on production servers. I check how your agent is exposed, close it up, and hand you a plain-English report.

What I do:

• Find what is exposed: open ports, the agent's gateway and admin UI, tokens in config files
• Update the agent and the OS to current versions
• SSH keys only, no root login, firewall and fail2ban
• Put the agent behind a reverse proxy with TLS and a login, or behind a private tunnel
• Limit tool permissions so a bad web page has far less it can make the agent do
• Back up config and memory, then prove the restore works
• Report: what was wrong, what I changed, and what is left

I work with SSH key access you add for me and remove when we're done.