n8n

the problem I'm working on with Ota. It's an open-source CLI for repo readiness that uses an `ota.yaml` contract to make setup, readiness, execution, and agent-safe guidance explicit for humans, CI, automation, and agents. For a monorepo, I'd think less in terms of "make the whole repo AI-ready" and more in terms of "make each contribution path explicit enough that an agent doesn't have to guess." We've also pressure-tested Ota on the `n8n` repo here: https://github.com/n8n-io/n8n/pull/30714. The PR is closed, but the proposed contract is still there to inspect. If you want to pressure-test your repo, share a simplified version of your setup and I can help think through what an Ota contract for it might look like. Ota: https://github.com/ota-run/ota Ota Examples: https://github.com/ota-run/examples

l your product make the project not open-source? Part of the point of Open Source is that the software isn't completely tied to a single company. If software is under one of those more restrictive licenses and the company goes under, the software is dead. If software under an open source license is developed by a company that goes under, one or even many other companies can continue working on it. This also applies while the company is alive, too; as you note, commercial companies developing open source software is a good thing, and preventing parallel development or forks is bad for the ecosystem.

If there's an existing business solving their issues and either it's not substantially n8n, and the n8n part isn't a product, service, or module being sold, it could be OK. This reads like if the complex functionality of your solution is being executed in connected systems external to n8n that it helps route through, the weighing of substantial is clearly outside. "Our license restricts use to "internal business purposes". In practice this means all use is allowed unless you are selling a product, service, or module in which the value derives entirely or substantially from n8n functionality. " --- Small workloads - You are correct in if it was an n8n process being packed and sold, even at small workloads, it could be a licensing issue. Conversely, if the small workload wasn't a process being sold to users. Commercial quoting - Makes sense, again. If I want to write a workflow in n8n and just skin it, i'm substantially using their software. To me, n8n can't do everything I need so I can't imagine a scenario where that I'd hit that, but absolutely other use cases could. Additionally, the commercial license seems to be for enterprise customers - maybe an option for small business/startup (headcount, or revenue) could be available. Seems reasonable enough, and have seen it with other tools. If substantial n8n use generated that kind of profit, I'd be happily paying for it to make sure it sticks around and is supported. It's like having an entire n8n dev team to keep n8n running for less than the cost of a developer. But I'm not there. For me, it's rare that I would rely solely on one tool that heavily. Having base inputs (especially for AI) in text form and feeding it up so different layers could be refactored to other tools over time (such as n8n's orchestration), there can be scenarios where the work is already substantially outside of n8n. At the same time, I look at tools like n8n for what they can do, using them as is, out of the box only, and not modifying them at all, because it then can receive updates. ---

is actually a benefit in my case since I mostly run n8n locally through docker. Whenever I spin it up, it auto catches up on new data it hasn't yet processed. n8n's AI nodes are first rate and more intuitive IMO than others. You can easily extend it with custom LangChain nodes as well if you're self hosting. If you want to play around with n8n locally, this tool makes it easy: https://github.com/LLemonStack/llemonstack/ I created it to make it easier for me to spin up and down project stacks on the same machine. Or see n8n's official repo for instructions on running locally: https://github.com/n8n-io/n8n