Skip to main content

Documentation Index

Fetch the complete documentation index at: https://conductor.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

An MCP tool is a named, specific action the AI is allowed to perform through the MCP connection—for example, fetching a particular kind of report, looking up structured data, or running a defined step that Conductor supports. MCP servers are like smart, AI assistants that have a menu of approved capabilities, not a free pass to “do anything on the internet.” Each item on that menu is a tool. When you ask something in natural language, the assistant decides which tool (if any) fits your request, calls it in a standard way, and then explains the result back to you in everyday language. That’s why tools feel powerful but still bounded: they are designed to do concrete jobs reliably, without improvising beyond limits.

​
Why “tools” matter for non-technical users

Without tools, a general-purpose AI only has public training knowledge. It cannot see your Conductor accounts, your tracked topics and prompts, or your performance metrics unless something bridges that gap. Tools let you cross that bridge with small, understandable steps. You still speak naturally; the system maps your intent to the right step (or a short sequence of steps). The complexity lives in how each tool is built and secured, not in what you have to type.

​
How this relates to Conductor

In Conductor’s platform, tools typically align with real product capabilities: the kinds of questions marketers and strategists already care about—visibility, performance, how brands show up in AI answers—grounded in your organization’s data. So when someone says “MCP server,” you can picture a set of tools an LLM may use on your behalf—always within Conductor’s and your company’s rules—so natural language can lead to serious work without you touching the underlying machinery.