Glossary
Stackmint glossary
Plain definitions of the terms you meet when you set up an AI operating system and build on the Stackmint platform.
Search docsBrowse the docs
Buyer terms
What you get with every OS
Each term links to the page that explains it in full.
Platform terms
How the platform is built
Under the hood, specialists are agents (Dendrites) that run workflows (Branches) made of steps (Buds).
Governance and control
Governed AI Capability
A governed AI capability is a packaged AI workflow tied to a specific business outcome, controlled by human approval gates, audit logs, scoped permissions, budget limits, and policy checks.
Read definitiongovernance controlHuman-in-the-Loop Gate
A human-in-the-loop gate pauses an AI workflow until an approved reviewer confirms, edits, or rejects the next action.
Read definitiongovernance architectureExecution Control Plane
An execution control plane manages approvals, audit logs, permissions, model routing, budgets, memory, and policy checks for AI workflows.
Read definitionRuntime architecture
Bud
A Bud is a reusable Stackmint primitive for logic, UI, connectors, prompt instructions, or review steps.
Read definitionworkflow primitiveBranch
A Branch is a complete Stackmint workflow that connects Buds into a governed execution path.
Read definitionrouting primitiveDendrite
A Dendrite is a Stackmint routing layer that chooses the right workflow, tool, model, or execution path.
Read definitionexecution surfaceHeadless AI Workflow
A headless AI workflow runs without a dedicated user interface, triggered by APIs, schedules, webhooks, Slack, CRM, or background jobs.
Read definitionprotocol interfaceMCP Tool
An MCP tool exposes a workflow or capability to agent-ready environments using the Model Context Protocol.
Read definitionWorkspace and client delivery
Client Workspace
A client workspace is a secure Stackmint environment where a partner deploys and manages capabilities for a specific client.
Read definitionplatform architectureCapability Cloud
A capability cloud is the execution platform used to build, deploy, govern, and monetize AI capabilities across clients.
Read definitionCommercial model
AI-Native Agency
An AI-native agency uses governed AI capabilities to operate client workflows, expand delivery capacity, and create recurring AI-powered services.
Read definitionmonetization infrastructureRevenue Rails
Revenue rails are the pricing, billing, usage metering, credit, revenue split, and payout systems that let partners monetize AI capabilities.
Read definition