What is tool calling?
Tool calling is how an AI agent takes action instead of only producing text. The model is given a set of tools, like search, a database query, or sending a message, and decides when to call one, with what inputs. The tool runs, returns a result, and the agent continues.
Tool calling is the mechanism that turns a language model from a text generator into something that gets work done. Without tools, a model can only talk about a task. With tools, it can actually do it.
How tool calling works
- The agent is given a list of tools, each with a description and expected inputs.
- For the current step, the model decides whether a tool is needed and which one.
- It calls the tool with specific inputs; the tool runs and returns a result.
- The agent reads the result and continues, often calling more tools until the task is done.
Standards like the Model Context Protocol make these tools reusable across agents and vendors, so you connect a system once.
Tool calling in Oido
Oido gives agents built-in tools plus any tool you connect through integrations and MCP. You control which tools each AI employee can use, and agents call them automatically as a task requires, on whichever model provider you run.
Frequently asked
Tool calling is the agent deciding to use a tool. MCP is a standard way to connect those tools so any compatible agent can use them. Tool calling is the action; MCP is the plumbing that makes tools reusable across agents and providers.
Each tool comes with a name and description of what it does and what inputs it needs. The model reads these and picks the right one for the current step, the same way a person reads a menu of options.
Yes, if you allow it. Tools can be read-only (look something up) or able to act (update a record, send an email). Good platforms let you control which tools an agent can use and add approval steps.