Lansweeper Workflow Assistant

Helps you build, fix and understand Lansweeper Flow Builder workflows: triggers, steps, filters, code blocks, loops and branches. Answers come from the Flow Builder documentation, not guessed from memory.

The Lansweeper Workflow Assistant skill covers the whole Flow Builder workflow lifecycle in one place. Describe what you want to automate, and it lays out the workflow step by step: trigger, then each connector action with its key inputs, then how each step’s output feeds the next. It writes ready-to-paste Lansweeper connector filters, list filters and code-block JavaScript, and works out which of three kinds of “filter” you mean (a Lansweeper data filter, a filter on a list already in the workflow, or a branch condition) before writing one. It uses a bundled copy of the Lansweeper connector reference (every action, field path and filter path), the core Flow Builder docs, about 25 built-in helper connectors and an index of every third-party connector. For apps like ServiceNow, Jira, Slack or Intune, it looks up the connector’s live documentation through the Lansweeper MCP’s docs search when connected, or opens the docs page directly when not, so you get the exact current actions and inputs. When something isn’t documented, such as a trigger payload shape, it says so and tells you how to check it (Listening mode, a test run) rather than inventing an answer.

You can even export your work-in-progress workflows and let your AI assistant adjust them for you before importing them again!

image 207

Why a Workflow Assistant Matters Here

  • IT/automation builder perspective

    “Which action do I use, what does the filter look like, and why does my code block return undefined?”

    • Ready-to-paste output: Lansweeper connector filters in the right shape for each action, list filter functions and code blocks that read previous step results correctly.
    • Wiring help: which outputs feed straight into the next action, when you need to map to plain asset keys, how to page through large estates, and why a step referenced after a branch might be empty.
    • Debugging: common failures (typed paths treated as literal text, renamed steps breaking code, a trigger change not saved, a workflow tested but never enabled) are diagnosed and fixed with the corrected input.
  • IT operations/governance perspective

    “Will this automation do what I think it does, safely, and keep running?”

    • Safe by default: destructive steps (Delete Assets, bulk Edit Assets, Delete IP Location or Asset Group) are flagged, with a suggestion to test with Max Records or a narrow filter first.
    • Runs in production: covers versioning and enabling (Save as new version), error handling and retries, and runner limits (execution time, memory, webhook payload size, sync timeouts, concurrency) before they cause a failure.
    • Grounding: action names, keys, filter paths and operators come from the documentation, so admins don’t end up debugging a step that never existed.

This skill turns Flow Builder from “read the docs and trial-and-error” into “describe the outcome and get a working workflow”, for people who have never written a filter or a line of JavaScript.

Requirements

Lansweeper MCP

Recommended, not required: https://mcp.lansweeper.com/mcp

When connected, the skill uses the MCP’s documentation search to pull live third-party connector pages (actions, inputs, authentication setup). Without it, the skill uses its bundled documentation and opens docs.lansweeper.com pages directly, which needs web access.

Lansweeper plan

Pro, Enterprise

LLM

Claude 3.5+, ChatGPT, Gemini (Antigravity), Copilot

Step by step setup guide

Step 1: Download
Download the skill package for your platform. Select Claude, ChatGPT, Gemini, or Copilot and the corresponding package downloads automatically.

Step 2: Install

  • Claude: install as a skill in Claude Desktop, Cowork or Code. The package includes SKILL.md plus six reference files; keep the folder structure intact. Optionally connect the Lansweeper MCP for live connector documentation.
  • ChatGPT: install as an app.
  • Gemini: load it into your agent’s skill folder. The Lansweeper MCP is optional.
  • Copilot: import it as a solution in Power Platform / Copilot Studio, with the reference files attached as the knowledge base.

Step 3: Ask

  • Claude: “Build me a workflow that adds every Windows device without CrowdStrike to an asset group called Missing EDR.”
  • ChatGPT: “Write a filter for List of Assets: Windows or Linux servers not scanned in the last 30 days.”
  • Gemini: “My code block says stepResults.listOfAssets is undefined — here’s the code, what’s wrong?”
  • Copilot: “Create a ServiceNow incident for every asset affected by a critical vulnerability.”
Related Skills
See all skills