Miqro Developer Docs

Welcome to the engineering documentation for the Miqro platform — the architecture, APIs, connectors, and integration surfaces behind miqromanage.app and miqrodesign.com.

These docs are for developers and technical partners who want to understand how the platform is put together, how its pieces talk to each other, and how to integrate with it.

Start here

• Architecture overview — the whole system on one page: the hub-and-spoke domain model, the AI agent fleet, and the intake → build → run lifecycle.
• Domains & SSO — how the backbone domain, the dashboards, the marketing sites, and single sign-on fit together.
• Multi-tenancy — how one app serves the operator and every customer, scoped by identity.
• The agent fleet — the agent execution model, the action queue, and the approval gate.

Tools

• The miqro CLI — the human/scriptable face of the platform: the same commands an agent calls via MCP, in your terminal and CI. CLI and MCP are two faces of one capability core.

Reference

• API overview — auth model, base URLs, and conventions for the HTTP API.
• API reference — the live, public-facing endpoint reference, generated from the running service.
• Connectors — every integration the AI team can connect to, generated live from each connector's capability manifest.
• MCP — where Model Context Protocol fits: the two directions (Miqro as client vs. server), today and on the roadmap.
• Agentic control via MCP 🔒 — how an agent drives external systems through a self-hosted MCP client without ever making an unapproved write (Enterprise / miqromanage).
• Coding sources 🔒 — the first high-value MCP targets: GitHub, trackers, and coding-tool servers.

These docs describe the platform's public and architectural surface. Internal operator-only endpoints (agent execution, internal service-to-service calls) are intentionally not enumerated here.