{{template "shellTop" "tools"}}
enabling requires step-up — assert a passkey first.
{{if .Msg}}maven drafted these from acts she couldn't run. Fill the command (argv, space-separated) and enable. A row in an mcp: scope came from an MCP server and already knows what it calls — check the command, then enable.
| name | capability | scope | from utterance | enable as |
|---|---|---|---|---|
{{.Name}} | {{capability .}} | {{.Scope}} | {{.Utterance}} |
grouped by capability domain. The dotted id is scope.domain.action — the same shape Hexis speaks — and it is derived from the row, so it always describes what the command actually does.
{{.Prefix}} {{len .Tools}}| capability | name | command | risk | |
|---|---|---|---|---|
{{capability .}} | {{.Name}} | {{join .Cmd " "}} |
{{$r := risk .}}{{if eq $r "irreversible"}}irreversible{{else if eq $r "destructive"}}destructive{{else}}safe{{end}} |
servers she connects OUT to. Their tools appear above as proposals — a configured server is a place she may look, not a capability she has. A stdio target is a process on this box; an http one on a loopback or LAN address is inside the network, so treat its tools accordingly.
| name | transport | target | state | tools |
|---|---|---|---|---|
{{.Name}} | {{.Transport}} | {{.Target}} |
{{if .Connected}}connected{{if .Server}} — {{.Server}}{{end}}{{else}}down{{if .Err}} — {{.Err}}{{end}}{{end}} | {{.Tools}} |
mcp.servers block to mavend.json to let her use an external tool server