vocabulary: 0.1
info:
  title: Continue Vocabulary
  description: Controlled vocabulary for Continue — the open-source AI code assistant for VS Code and JetBrains, plus the Continue Hub IDE API and Continue CLI.
  created: '2026-05-25'
  modified: '2026-05-25'
terms:
  - id: assistant
    label: Assistant
    definition: A YAML-defined Continue Hub configuration bundling models, rules, prompts, docs, MCP servers, and context providers that an IDE extension or CLI consumes.
    aliases: [continue-assistant, hub-assistant]
  - id: block
    label: Block
    definition: A reusable Continue Hub building block — a single model, rule, prompt, doc, MCP server, or context provider — that can be composed into an assistant.
  - id: hub
    label: Continue Hub
    definition: Continue's hosted registry at hub.continue.dev where teams publish and share assistants, blocks, models, rules, prompts, docs, and MCP servers.
    aliases: [continue-hub]
  - id: rule
    label: Rule
    definition: A natural-language guideline an assistant applies to every chat, edit, or autocomplete interaction.
  - id: prompt
    label: Prompt
    definition: A reusable, named instruction (often shipped in a .prompt file) the IDE or CLI can invoke.
  - id: check
    label: Check
    definition: A markdown-defined Continue agent stored in .continue/checks/ that runs in CI on every pull request and surfaces as a GitHub status check.
  - id: agent
    label: Agent
    definition: A Continue execution mode in which the model autonomously edits files, runs commands, and uses tools to complete a task.
  - id: ide-extension
    label: IDE Extension
    definition: The open-source Continue plugin shipped for VS Code and JetBrains IDEs.
  - id: cli
    label: Continue CLI
    definition: The open-source Continue command-line interface that runs checks locally and in CI.
  - id: mcp-server
    label: MCP Server
    definition: A Model Context Protocol server an assistant connects to in order to expose extra tools, prompts, or resources.
  - id: context-provider
    label: Context Provider
    definition: A Continue plugin that injects additional information (code, docs, terminal output, git diffs) into the model's context window.
  - id: model-role
    label: Model Role
    definition: The role an LLM plays inside a Continue assistant — chat, edit, apply, autocomplete, embed, or rerank.
  - id: free-trial
    label: Free Trial
    definition: Continue's hosted, quota-limited trial of chat and autocomplete that lets new users try the IDE without configuring an LLM provider.
  - id: models-add-on
    label: Models Add-On
    definition: A paid subscription, billed via Stripe through /ide/get-models-add-on-checkout-url, that grants additional Continue-managed model usage on top of a plan.
  - id: on-prem-proxy
    label: On-Prem Proxy
    definition: A self-hosted Continue proxy organizations route model traffic through to keep prompts inside their own network perimeter.
  - id: organization
    label: Organization
    definition: A team scope on Continue Hub that owns assistants, blocks, secrets, and policy.
  - id: policy
    label: Policy
    definition: Organization-level Continue settings — allowed models, secret handling, telemetry, and proxy enforcement — returned by GET /ide/policy.
  - id: mission-control
    label: Mission Control
    definition: Continue's metrics and observability surface for tracking checks, agent runs, and outcomes across an organization's repositories.
  - id: continuous-ai
    label: Continuous AI
    definition: Continue's positioning — source-controlled AI standards enforced on every pull request, "standards as checks, enforced by AI, decided by humans."