Skip to main content
Rilo doesn’t need special syntax or commands. It understands plain language — but the quality of what it delivers depends heavily on how clearly you describe what you need.

The core principle: be specific about the outcome

Rilo will make reasonable choices wherever you leave things unspecified. That’s useful when the details don’t matter to you, but if they do matter, spell them out.
Weak:
Rilo will produce something, but you’ll get a general overview when you might have wanted something much more specific.Better:
Naming the context (“renewal call”), specifying the sources (“LinkedIn is fine”), and describing the format (“short briefing, 2 minutes”) all sharpen the result.

Provide context Rilo can’t see

Rilo can access your connected tools, but it doesn’t know things that live only in your head. If the task involves a specific situation, relationship, or background — include it:
Background context turns a generic output into something useful.

Specify format when it matters

If you’re going to paste results into a doc, send them to someone, or present them — say so:
  • summarize as 5 bullet points
  • put it in a table with columns for: name, title, email, LinkedIn
  • format as a Slack message, keep it conversational
  • write it up as a proper report with headings
  • give me just the numbers, no explanation needed

Tell Rilo what tools to use (or avoid)

If you have a preference for which tool Rilo should use for a task, say so:

Ask for clarification up front

If you’re not sure whether you’ve given Rilo enough to go on, you can ask it to check before starting:

What Rilo will do when a task is ambiguous

If the task is too ambiguous to make a reasonable interpretation, Rilo will ask a clarifying question before proceeding. It won’t guess when getting it wrong would waste significant work. For shorter tasks, Rilo may make a reasonable assumption and proceed, then note what it assumed in the result.