use-alpic

name	use-alpic
description	Use Alpic Cloud and the `alpic` CLI for deployed MCP apps/servers, including deploys, builds, logs, debugging, environments, environment variables, tunnels, playgrounds, audits, auth, domains, IP restrictions, analytics, insights, versioning, and MCP Registry publishing.

Alpic Cloud

Alpic hosts and operates MCP apps and servers. Use this skill for deployment, runtime operations, distribution, security, observability, and publishing on Alpic Cloud.

The reference files contain the task workflows and command details. Load only the file(s) relevant to the user's request.

When To Use

• The user mentions Alpic Cloud, alpic, Alpic deployments, Alpic logs, Alpic playground, Alpic tunnels, Alpic environments, or Alpic publishing.
• The user wants to operate an already-built MCP app/server.
• The user wants production/distribution guidance after building an MCP app.

Use build-mcp-apps instead when the work is primarily about designing or implementing the MCP app/server itself.

References

Deploy and operate:

• Deploy an MCP app/server: deploy.md
• Build configuration, runtimes, start commands: builds.md
• Failed deploys and startup failures: debug-deployment.md
• Runtime and deployment logs: logs.md
• Environment variables: environment-variable.md
• Environments, branches, staging/preview/prod: environments.md
• GitHub auto-deploys: git.md
• Local HTTPS tunnels: tunnel.md

Inspect and test:

• Usage metrics and project analytics: analytics.md
• Readiness checks and launch audits: audit.md
• Playground configuration and headers: playground.md
• Playground example prompts: example-prompts.md
• Public URLs and endpoint shapes: endpoints.md

Secure and distribute:

• OAuth, DCR proxy, API-key protection: auth.md
• Fixed outbound IP and inbound IP whitelisting: network-security.md
• Custom domains and DNS validation: domains.md
• User intents and feedback capture: insights.md
• Host caching and submitted-version strategy: versioning.md
• MCP Registry publishing: publish.md

Interface Choice

• Prefer the alpic CLI for deploys, logs, environment variables, environments, git, tunnels, audits, playground settings, and publishing.
• For agent-driven CLI work, prefer non-interactive commands after discovering and confirming the required team, project, and environment IDs. Use interactive prompts only when the user is intentionally driving the selection or the CLI offers no equivalent flags.
• Use the Alpic dashboard for custom domains, DCR client pools, fixed outbound IP, inbound IP whitelisting, visual analytics, and Beacon rendering checks.
• Use the hosted Alpic MCP server/app for agent-driven project browsing, logs, and analytics when available.
• Use REST APIs only for automation or gaps in CLI/MCP coverage.

Guardrails

• Store secrets in Alpic environment variables, not source files.
• Discover and confirm the target team, project, and environment before a first deployment or before changing production config, auth, domains, IP restrictions, fixed outbound IP, or environment variables. Do not treat an interactive picker as user confirmation when multiple plausible targets exist.
• Use build/deployment logs for deploy failures and runtime logs for live behavior.
• Re-deploy after changing environment variables.
• Do not delete DCR client pools unless intentionally breaking existing client registrations.
• Publish only after a production deploy plus audit/playground verification.