MCP tool
Callable function exposed by an MCP server with a name, description, and input schema. The hosted Intelligence server offers list_capabilities, analyze_compensation, analyze_skills, analyze_market, discover_roles_and_cities, and find_adjacent_opportunities, each returning a structured answer.
In plain language
An agent discovers the available functions, then calls one with arguments matching its schema.
In a sentence
- The MCP tool described its accepted arguments before the agent called it.
- We used the MCP tool to discover role and city identifiers first.
Usage note
A tool's schema describes its inputs. The caller's plan and scopes still determine whether the hosted server allows that particular call.
See also
- MCP
MCP, the Model Context Protocol, is an open standard that lets AI agents call external tools and data through a JSON-RPC interface.
- Decision-ready response
A decision-ready response is the output every Orbyt MCP tool returns: a one-sentence answer, data split into primary and supporting, context (methodology version, sample size, confidence and source breakdown), a citation (methodology URL, data_point_id, request_id), follow_up_suggestions, limitations and an optional cache_ttl_seconds.
- JSON-RPC 2.0
Remote procedure call protocol whose requests and responses are JSON objects.
- Scope
A scope is a permission on an API key that decides which endpoints the key may call, such as intelligence:read, compensation:read, skills:read, market:read or intelligence:lineage.
Cite this term.
CC BY 4.0, attribution required.
Orbyt Labs, MCP tool (Developer Glossary), Nov 6, 2026. https://
Orbyt Labs Glossaries (Nov 14, 2026), https://
Last updated .