Glossary
Glossary
| Term | Meaning |
|---|---|
| Team | Administrative group that owns projects and billing settings. |
| Project | Backend data model, data, API, membership, and service configuration. |
| Studio | The Flexkit app, as a desktop app or at flexkit.app; it renders each project’s deployed schema. Developers can also self-host it. |
| Schema | The project’s data model: entities, attributes, relationships, scopes, and spaces. Built with the agent or defined in code. |
| Entity type | Definition of a kind of record, such as Product. |
| Record | One saved instance of an entity type. |
| Attribute | A field in an entity definition. |
| Relationship | A link between records, optionally with relationship-scoped attributes. |
| Scope | A named data variant, such as a locale; not an access grant. |
| Space | Permission boundary assigned to resources and granted to members. |
| Role | A subject’s permitted level of access within a team or project. |
| Extension | Behaviour added through skills, plugins, custom tools, automations, or the API; Studio UI is not extended from code. |
| Plugin | A catalog package that connects an external service’s tools and skills to the agent; see plugins. |
| Integration | A connection to an external product or service; see integrations. |
| Knowledge | Documents added to a project that the agent searches and cites; see Knowledge. |
| Package | An npm distribution unit, such as @flexkit/studio or @flexkit/schema. |
| App | A built-in Studio workspace (AI, Desk, Asset Manager, Knowledge, Graph, Explorer), selected in Studio’s app rail. |
| API token | A credential for authenticated requests; keep it out of public client code. |
| Skill | Instructions available to AI, written in Studio, proposed by the agent, or defined in code. |
| Custom tool | Server-side code callable by AI through the configured handler. |
| Automation | Saved AI instructions, policy, attachments, and execution configuration. |
| Trigger | An event, schedule, or incoming webhook that starts an automation. |
| Run | An execution with its own status, output, and diagnostic history. |
| Approval | A decision about a proposed mutation or plugin call. |
| AI credits | Team credit used for AI execution; distinct from infrastructure usage. |
See core concepts for the architecture and Dashboard members for access. CLI authentication may also use the word scope; that does not refer to schema data variants.
Last updated on