Model Context Protocol, or MCP, lets AI coding assistants such as Claude Code, Cursor, and other MCP-compatible clients call external APIs in a structured way. With Qlerify’s MCP server and plugin skills, your assistant can read, create, and update workflows and domain models directly in Qlerify.
Create your API token in the Qlerify UI:
Next, add the Qlerify MCP server in your client’s MCP settings. This example shows how to add it with Claude:
> claude mcp add --transport http qlerify https://mcp.qlerify.com \
--header "x-api-key: YOUR_API_KEY" Your AI assistant can now connect to Qlerify to read and create workflows and domain models.
The official bundle lives at github.com/Qlerify/qlerify-plugins. It includes skills that teach the assistant how to use the MCP tools effectively: what order to run steps in, how to name things, and how to go from an idea to a modeled workflow.
To install with Claude, start Claude from the command line, then run the following two commands:
/plugin marketplace add qlerify/qlerify-plugins
/plugin install mcp-companion@qlerify-plugins
Find setup instructions for other clients, such as Cursor, Lovable, VS Code, and Gemini CLI, under User Profile → Integrations in Qlerify.
extract-aggregate (the extractor) — Use this when you’re starting from existing code. It walks the assistant through isolating one DDD aggregate at a time: root entity, related entities, value objects, commands, domain events, read models, invariants, and external references. The output is structured so you can review it and then feed it into modeling in Qlerify. Natural prompts: “reverse engineer the Order module,” “extract the Subscription aggregate from this repo,” “document this legacy service as a domain model.”workflow-creation — Use this when you’re building or extending a workflow inside Qlerify: swimlanes, groups, domain events, entities, commands, read models, cards, bounded contexts. It follows a clear sequence (e.g. lanes → groups → events → entities → commands → read models → bounded contexts) so the diagram stays consistent. Natural prompts: “create a workflow for checkout,” “add domain events and commands for orders,” “set up bounded contexts for this process.”
Often, you use extract-aggregate on code or documentation to create a clean aggregate description. Then, you use workflow-creation to implement that structure in Qlerify.
For greenfield design, you can skip extraction and go straight to workflow creation.
