Query and operate Port.io — the Internal Developer Portal that governs your SDLC. Use when working with software catalogs, blueprints, entities, scorecards, self-service actions, incidents, or DevOps/platform requests. Triggers on mentions of Port, port.io, IDP, developer portal, scorecards, blueprints, service catalog, or self-service actions.
설치
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
Query and operate Port.io — the Internal Developer Portal that governs your SDLC. Use when working with software catalogs, blueprints, entities, scorecards, self-service actions, incidents, or DevOps/platform requests. Triggers on mentions of Port, port.io, IDP, developer portal, scorecards, blueprints, service catalog, or self-service actions.
Use the Port MCP tools automatically when the user works with their Internal Developer Portal — the software catalog, scorecards, self-service actions, or any SDLC governance task on Port.io.
Regions
Port operates in two regions, exposed as two MCP servers in this plugin:
Region
MCP server
Endpoint
US
port-us
https://mcp.us.port.io/v1
EU
port-eu
https://mcp.port.io/v1
Use one region — whichever matches the user's Port account. If unsure which region they are on, call describe_user_details on each and use the one that authenticates, or ask. Do not duplicate calls across both regions.
When to Use
Catalog discovery: list blueprints/entities, find services, owners, dependencies.
Scorecards & quality: check scorecard rules, levels, and which entities pass/fail.
Self-service: list and run self-service actions, then track their run status.
Knowledge: answer questions about the org's software using search_port_knowledge_sources.
Portal config: inspect the sidebar, manage folders and dashboard widgets.
How to Use
Start with describe_user_details to confirm the active org/region when context is missing.
Use list_blueprints → list_entities to explore the catalog before answering catalog questions.
Use list_scorecards for quality/compliance questions.
For self-service: list_actions to find the action, run_action to execute, then track_action_run to monitor until completion. Confirm with the user before running any action that creates or changes infrastructure.
For org-knowledge questions, prefer search_port_knowledge_sources over guessing.
Important Notes
Authentication is handled by the Port MCP server's OAuth flow (/mcp to connect). No API key env var is required.
run_action may trigger real infrastructure changes — treat it as a write operation and confirm intent first.
Some MCP servers expose additional capabilities via load_skill; call it to discover Port-specific guided workflows when available.
