Skip to main content
Glama

deployment_trigger

Initiate deployments for services on Railway-MCP by specifying project, service, environment, and commit SHA. Use for code changes, configuration updates, or rollbacks. Requires service_list for prerequisites.

Instructions

[API] Trigger a new deployment for a service

⚡️ Best for: ✓ Deploying code changes ✓ Applying configuration updates ✓ Rolling back to previous states

⚠️ Not for: × Restarting services (use service_restart) × Updating service config (use service_update) × Database changes

→ Prerequisites: service_list

→ Alternatives: service_restart

→ Next steps: deployment_logs, deployment_status

→ Related: variable_set, service_update

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
commitShaYesSpecific commit SHA from the Git repository
environmentIdYesID of the environment
projectIdYesID of the project
serviceIdYesID of the service

Implementation Reference

  • The handler function for the 'deployment_trigger' tool. It receives the input parameters and calls the deploymentService to initiate the deployment.
    async ({ projectId, serviceId, environmentId, commitSha }) => {
      return deploymentService.triggerDeployment(projectId, serviceId, environmentId, commitSha);
    }
  • Zod input schema defining the parameters required for triggering a deployment.
    {
      projectId: z.string().describe("ID of the project"),
      serviceId: z.string().describe("ID of the service"),
      environmentId: z.string().describe("ID of the environment"),
      commitSha: z.string().describe("Specific commit SHA from the Git repository")
    },
  • Registers all tools, including 'deployment_trigger' from deploymentTools, to the MCP server.
    export function registerAllTools(server: McpServer) {
      // Collect all tools
      const allTools = [
        ...databaseTools,
        ...deploymentTools,
        ...domainTools,
        ...projectTools,
        ...serviceTools,
        ...tcpProxyTools,
        ...variableTools,
        ...configTools,
        ...volumeTools,
        ...templateTools,
      ] as Tool[];
    
      // Register each tool with the server
      allTools.forEach((tool) => {
        server.tool(
          ...tool
        );
      });
    } 
  • Helper service method called by the tool handler to trigger the deployment via the API client, includes rate limiting wait.
    async triggerDeployment(projectId: string, serviceId: string, environmentId: string, commitSha?: string) {
      try {
        // Wait for 5 seconds before triggering deployment
        // Seems like the LLMs like to call this function multiple times in combination
        // with the health check function and the list deployments function
        // so we need to wait a bit to avoid rate limiting
        await new Promise(resolve => setTimeout(resolve, 5000));
        const deploymentId = await this.client.deployments.triggerDeployment({
          serviceId,
          environmentId,
          commitSha
        });
    
        return createSuccessResponse({
          text: `Triggered new deployment (ID: ${deploymentId})`,
          data: { deploymentId }
        });
      } catch (error) {
        return createErrorResponse(`Error triggering deployment: ${formatError(error)}`);
      }
    }
  • Repository helper that performs the GraphQL mutation to trigger a new deployment on the Railway API.
    async triggerDeployment(input: DeploymentTriggerInput): Promise<string> {
      const { commitSha, environmentId, serviceId } = input;
      const data = await this.client.request<{ serviceInstanceDeployV2: string }>(`
        mutation serviceInstanceDeployV2($commitSha: String, $environmentId: String!, $serviceId: String!) {
          serviceInstanceDeployV2(
            commitSha: $commitSha
            environmentId: $environmentId
            serviceId: $serviceId
          )
        }
      `, { commitSha, environmentId, serviceId });
    
      return data.serviceInstanceDeployV2;
    }

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/jason-tan-swe/railway-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server