Tool reference

Every tool the Rius MCP server exposes, with arguments and example invocations. All tools are read-only and return Markdown, which your client renders as tables and summaries.

Two conventions apply across the trace tools:

workspace is optional everywhere: pass a workspace name or ID, or omit it to use your default workspace. Discover workspaces with list_workspaces.
hours is a lookback window ending now (default 24). “This week” is hours: 168.

You rarely invoke tools by hand. Ask your client a question, and it picks the tools and arguments itself. The examples show what the client sends, which helps when a question is not picking the tool you expect.

get_me

Your identity (user ID, external ID, email) and the workspaces you can access.

No arguments.

Who am I in Rius?


{ "name": "get_me", "arguments": {} }

list_workspaces

The workspaces you have access to, as an ID and name table. Use it to find the right value for the workspace argument of the trace tools.

No arguments.

Which Rius workspaces do I have?


{ "name": "list_workspaces", "arguments": {} }

agent_traces_summary

Summary KPIs for agent traces over a time window: trace count, LLM calls, error rate, total and average cost, token usage, and latency percentiles (p50/p95/p99).

Argument	Type	Default	Description
`workspace`	string	your default	Workspace name or ID
`hours`	int	`24`	Lookback window in hours
`service`	string	all services	Only traces involving this service name

How did my agents do over the last 24 hours?


{ "name": "agent_traces_summary", "arguments": { "hours": 24 } }

Summarize the checkout-agent service this week.


{
  "name": "agent_traces_summary",
  "arguments": { "hours": 168, "service": "checkout-agent" }
}

list_agent_traces

Recent agent traces, one row per trace, newest first: trace ID, root span name, service, start time, duration, span count, status, model, tokens, and cost.

Argument	Type	Default	Description
`workspace`	string	your default	Workspace name or ID
`hours`	int	`24`	Lookback window in hours
`service`	string	all services	Only traces involving this service name
`status`	string	all	`Ok` or `Error`
`limit`	int	`20`	Maximum rows returned

Show me the failed runs from the last two days.


{
  "name": "list_agent_traces",
  "arguments": { "hours": 48, "status": "Error" }
}

The trace_id column feeds get_agent_trace. When more traces exist than limit, the reply says so; narrow the window or add a filter rather than raising the limit past what a chat can usefully display.

get_agent_trace

One trace’s span waterfall: every span in time order, indented by depth, with duration, status, and the LLM details (model, tokens, cost) where present. This is where an investigation ends up, since it shows which step failed or where the time went.

Argument	Type	Default	Description
`trace_id`	string	required	The trace ID, from `list_agent_traces`
`workspace`	string	your default	Workspace name or ID

Break down trace 4bf92f3577b34da6a3ce929d0e0e4736 for me.


{
  "name": "get_agent_trace",
  "arguments": { "trace_id": "4bf92f3577b34da6a3ce929d0e0e4736" }
}

Next steps

Example workflows: the three tools composed into complete investigations.
Client setup: connecting a client if you have not yet.