한가경

Hi Lindy team, I'm curious about how non-technical teams use Lindy in practice. Since not everyone on a team is a developer, I'm wondering: 1. How do you recommend teams share and reuse Lindies across team members? - Is there a team library or template sharing feature? 2. What's your approach for teams where some members build automations and others just need to use them? - Can non-technical users trigger/run Lindies built by others? 3. Do you have any best practices for "automation handoff" - where the person who built it isn't the one maintaining it? We're running AI automation workshops in Korea, and many companies ask about scaling from individual automation to team-wide adoption. Would love to learn from your experience! Best, 가경

Hi team, I’m running into a recurring issue with automated actions in Lindy. Even after refreshing the browser session, the task keeps failing with the same error: “An error occurred. We detected excessive failed actions in this task and automatically stopped it.” I’ve attached screenshots showing the repeated failure points. It seems that Lindy is trying to perform steps like clicking the logout button or typing into the address bar, but the actions don’t execute correctly and the task is halted. Could you help me understand what’s causing this repeated error, and whether there’s anything I should adjust on my side (e.g., browser environment, task structure, or step configuration)? Thanks in advance for your support!

Hi Lindy team! I’ve been rebuilding the Brand Monitoring agent from scratch, and I’m running into an issue as the workflow becomes more complex. When I expand the logic (e.g., multiple search queries, sentiment sources, different channels to monitor), Lindy automatically creates many separate nodes — sometimes more than I expected — and several of them end up disconnected or duplicated. This makes the workflow harder to maintain and debug. Could you advise on: Best practices for keeping a complex workflow manageable in Lindy How to prevent node explosion when branching logic or parallel actions are needed Whether there’s a recommended pattern (e.g., grouping, custom functions, consolidating steps) for workflows that involve multiple research/parse/summarize steps How to identify when a workflow should be split into multiple tasks instead of one large agent Any guidance or examples for scalable workflow patterns would be greatly appreciated!

Hi Lindy team, I'm building an agent that retrieves a specific Google Spreadsheet cell and posts the value to a Slack channel both on a weekly schedule and on-demand. The scheduled trigger works well, but on-demand triggering required configuring a keyword-based Slack trigger, and even then the agent only works when I manually add a custom step prompt inside the workflow. The current issues are:

Could you advise how to optimize the agent for sub-second or near-instant (<1s) response time for simple spreadsheet-to-Slack retrieval? Any best practices for model selection, step design, or caching would be greatly appreciated. Thank you!