Governed execution for people and AI agents
Learn the access model, carry out console tasks, connect automation, and look up exact commands and settings.
# Request a time-bound work session,# then open a browser terminalalpacon work-session create \ --purpose "nginx fix" \ --scope websh --server web-01 \ --expires-in 2h --wait --usealpacon websh web-01# Automation authenticates with an API token# access is scoped by the token's server allowlistalpacon login --workspace myworkspace \ --region us1 -t $ALPACON_API_TOKENalpacon exec web-01 -- systemctl restart nginx# Connect an AI agent over MCP# (hosted, browser sign-in, no token)claude mcp add --transport http \ alpacon https://mcp.alpacon.io/mcp # Then ask in plain language:# "List servers in the production workspace"Browse the docs
Integrate
CI/CD, REST API, MCP, webhooks, and SSO.
- GitHub Actions
- Automate with the REST API
- Connect an AI agent with MCP
- Send notifications to a webhook
- Single sign-on (SSO)
Start with your role
For engineersCreate a workspace, register your first server, and connect from a browser terminal in minutes.QuickstartFor workspace adminsInvite your team, register servers, wire up group access, and set approval and security policies.Set up your workspaceFor approversRead a work session or token request, decide well, and follow it through the audit trail.The approver's guideFor AI agentsConnect an agent over MCP, scope it to a task, keep a human on approvals, and audit everything it did.Give AI agents governed access
How it works
One governed path for every action: request, approve, audit.
- 1
Request
Ask for the servers, features, and time window you need, with a stated purpose.
Work sessions - 2
Approve
Access rules determine whether an action is allowed, needs approval, or is refused.
Approvals - 3
Audit
Review session history, command records, and approval decisions.
Audit