The power of the CLI

7 min Intermediate

What you'll learn

  • Why running Landbase through an assistant opens up more than the web app alone
  • Connect another tool from your stack so the agent can work with it alongside Landbase
  • Chain several steps into a single workflow instead of running each by hand
  • Hand a finished workflow to the agent and let it carry out the steps
  • Set the permissions and guardrails that keep automated work under control

The CLI runs inside your AI assistant, which means Landbase becomes one capability among all the others you have connected there. That setup lets you move past building lists and into running a motion, with the agent searching, enriching, and then handing results straight to the tools where you work. This lesson shows how to connect those tools and chain the steps together into something that runs for you.

Prompt
Build a list of in-market accounts, dedupe it against my CRM, and write the new ones back

Landbase inside your assistant

Because the CLI runs inside your assistant, Landbase sits alongside the other tools you have already connected there. That lets you move past building one list at a time and into running a motion, where the agent searches, enriches, and passes results onward in a single flow.

Connect a tool from your stack

Connect another tool, such as your CRM, so the agent can read from it and write to it directly alongside Landbase. With the tools in one place, the agent can compare a new list against what you already have and add only what is missing.

Chain steps into a workflow

Several steps can be combined into one workflow that runs end to end, such as build a list, enrich it, dedupe it against your CRM, and write the result back. Rather than running each step by hand, you describe the sequence once.

Hand the workflow to the agent

Once the workflow is defined, hand it to the agent and let it carry out the steps in order. This is the shift from doing the work to directing it, where you set the goal and the agent executes the steps.

Permissions and guardrails

Automated work stays under your control through permissions and guardrails. You decide what the agent can read and write, and you can review what it did before you rely on the result. Start with read access and a review step, then widen control as you build trust.

Try it in Landbase

  1. Connect one tool you use every day, such as your CRM, to your assistant.
  2. Ask the agent to build a small Landbase list and write the results into that tool.
  3. Watch it carry out both steps in sequence.
  4. Review what it did before you rely on the result.
Tip
Start every automated workflow with read-only access and a review step. Widen the agent’s permissions only once you trust the output.