An agent engine with no SDK, just TOML and webhooks
I was looking for a way to add AI agents to multiple systems that I maintain.<p>I had a few requirements:<p>- I work in multiple programming languages, so it needs to work for all of them.<p>- I can control how tools access my system.<p>- I can override the agent loop to try diff
An engine for agents that live in the cloud instead of on a laptop. Connect them to Slack, give them MCPs, and control the loop from your own system.
Build agents your whole team can use substructure.ai Docs Pricing llms.txt GitHub Get started Build agents your whole team can use An engine for agents that live in the cloud instead of on a laptop. Connect them to Slack, give them MCPs, and control the loop from your own system. Get started Read the docs substructure.toml name = "example" [llm.openrouter] type = "openrouter" [mcp.notion] url = "https://mcp.notion.com/mcp" [mcp.linear] url = "https://mcp.linear.app/mcp" [agent.ops] llm = "openrouter" model = "deepseek/deepseek-v4-flash-073
An engine for agents that live in the cloud instead of on a laptop. Connect them to Slack, give them MCPs, and control the loop from your own system.
Build agents your whole team can use substructure.ai Docs Pricing llms.txt GitHub Get started Build agents your whole team can use An engine for agents that live in the cloud instead of on a laptop. Connect them to Slack, give them MCPs, and control the loop from your own system. Get started Read the docs substructure.toml name = "example" [llm.openrouter] type = "openrouter" [mcp.notion] url = "https://mcp.notion.com/mcp" [mcp.linear] url = "https://mcp.linear.app/mcp" [agent.ops] llm = "openrouter" model = "deepseek/deepseek-v4-flash-073
