Provides tools for querying the service catalog, fetching TechDocs, looking up ownership, exploring dependency graphs, and scaffolding new services from golden-path templates via Backstage.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcp-developer-portal search catalog for payment-service"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP server wrapping Backstage — query service catalog, fetch TechDocs, and scaffold services via AI agents.
mcp-developer-portal is a Model Context Protocol server that exposes your
Backstage internal developer portal to AI agents. It enables LLMs to query
the service catalog, fetch TechDocs documentation, look up ownership, explore
dependency graphs, and scaffold new services from golden-path templates.
Built for platform engineering teams in regulated financial services and manufacturing who use Backstage as their engineering system of record.
Related MCP server: jt-mcp-server
Tool | Description |
| Search the Backstage catalog by name, kind, or owner |
| Get full details for a catalog entity |
| Fetch TechDocs documentation for an entity |
| Look up who owns a service or component |
| Get upstream and downstream dependencies |
| Create a new service from a golden-path template |
| List available scaffolding templates |
pip install mcp-developer-portal
export BACKSTAGE_URL=https://your-backstage.example.com
export BACKSTAGE_TOKEN=your_backstage_token
mcp-developer-portalVariable | Description | Required |
| Backstage instance URL | Yes |
| Backstage API token | No |
| Industry context (fintech/manufacturing) | No |
See CONTRIBUTING.md.
Apache 2.0 — see LICENSE.
Get upstream and downstream dependencies for a service.
Inputs: name (required), kind. Returns: result
Get full details for a catalog entity.
Inputs: name (required), kind, namespace. Returns: result
Look up who owns a service or component.
Inputs: name (required), kind. Returns: result
Fetch TechDocs documentation for a catalog entity.
Inputs: name (required), kind, namespace. Returns: result
List available golden-path scaffolding templates.
Inputs: no named parameters. Returns: result
Scaffold a new service from a golden-path template.
Inputs: template_name (required), service_name (required), owner (required), description. Returns: result
Search the Backstage service catalog by name, kind, or owner.
Inputs: query, kind, owner. Returns: result
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 7 tools
Each tool has a uniquely defined purpose: dependencies, entity details, ownership, techdocs, templates, scaffolding, and catalog search. No two tools overlap in functionality.
All tool names follow a consistent verb_noun pattern (e.g., get_dependencies_tool, list_templates_tool, scaffold_service_tool), with verbs like get, list, search, and scaffold clearly indicating the action.
Seven tools is a well-scoped set for a developer portal. They cover key operations without being excessive or insufficient for the intended domain.
The tool surface covers essential read operations (get, list, search) and a key write operation (scaffold). Minor gaps exist, such as missing create/update for entities beyond scaffolding, but the core developer portal workflow is well-supported.
The Cortex MCP server provides read-only access to real-time engineering context from the Cortex developer portal, allowing AI coding assistants to answer natural language questions about your organization's catalog (microservices, libraries, domains, teams, infrastructure), scorecards (engineering standards and best practices), initiatives (goals and deadlines), and Engineering Intelligence metrics. It includes tools for querying documentation, tracking personal entities, and accessing AI-assisted insights across the entire Cortex ecosystem.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
Remote MCP server for supportsheep: run AI interviews and manage support content for your blog.
Remote MCP server for supportsheep: run AI interviews and manage support content for your blog.
