Render MCP Server (render-oss/render-mcp-server)

Render MCP Server

https://github.com/render-oss/render-mcp-server
Admin
The Render MCP Server is a Model Context Protocol (MCP) server that allows interaction with Render...

Tokens:10,082
Snippets:29
Trust Score:8.4
Update:5 months ago
Show doc for...
Context Summary (auto-generated)
Raw
# Render MCP Server

The Render MCP Server is a Model Context Protocol (MCP) implementation that provides programmatic access to Render cloud platform resources through LLM integrations. Built in Go, it exposes Render's infrastructure management capabilities including web services, static sites, PostgreSQL databases, Redis key-value stores, logs, metrics, and deployments through a standardized MCP interface.

This server enables AI assistants to create and manage cloud infrastructure, query databases, monitor application performance, troubleshoot issues through log analysis, and track deployment history. It supports both stdio transport for desktop AI applications and HTTP transport with session management for web-based integrations, making it suitable for interactive development workflows and automated infrastructure operations.

## APIs and Key Functions

### Workspace Management

List and select workspaces to scope all subsequent operations to the correct organizational context.

```go
// List all accessible workspaces
// Returns JSON array of workspace objects with id and name fields
// Automatically selects workspace if only one exists
import (
    "context"
    "github.com/render-oss/render-mcp-server/pkg/owner"
    "github.com/render-oss/render-mcp-server/pkg/client"
)

func listWorkspaces(ctx context.Context, c *client.ClientWithResponses) {
    ownerRepo := owner.NewRepo(c)
    workspaces, err := ownerRepo.ListOwners(ctx, owner.ListInput{})
    if err != nil {
        panic(err)
    }
    // Response: [{"id":"own-123abc","name":"My Team","email":"team@example.com"}]
}

// Select a specific workspace for all operations
// Required before creating resources
func selectWorkspace(ctx context.Context, ownerID string) {
    sess := session.FromContext(ctx)
    err := sess.SetWorkspace(ctx, ownerID)
    if err != nil {
        panic(err)
    }
    // All subsequent operations will use this workspace
}
```

### Web Service Management

Create and manage web services with automatic deployments from Git repositories.

```go
// Create a new Node.js web service
// Automatically deploys on branch updates
import (
    "context"
    "github.com/render-oss/render-mcp-server/pkg/service"
    "github.com/render-oss/render-mcp-server/pkg/client"
)

func createWebService(ctx context.Context, c *client.ClientWithResponses) {
    serviceRepo := service.NewRepo(c)

    // Build request with required parameters
    envVars := []client.EnvVarInput{}
    envVar := client.EnvVarInput{}
    envVar.FromEnvVarKeyValue(client.EnvVarKeyValue{
        Key:   "NODE_ENV",
        Value: "production",
    })
    envVars = append(envVars, envVar)

    repo := "https://github.com/username/my-app.git"
    branch := "main"
    region := client.Oregon
    plan := client.PaidPlanStarter

    nativeEnv := client.NativeEnvironmentDetailsPOST{
        BuildCommand: "npm install && npm run build",
        StartCommand: "npm start",
    }

    envDetails := client.EnvSpecificDetailsPOST{}
    envDetails.FromNativeEnvironmentDetailsPOST(nativeEnv)

    webServiceDetails := client.WebServiceDetailsPOST{
        Runtime:            client.ServiceRuntimeNode,
        EnvSpecificDetails: &envDetails,
        Plan:               &plan,
        Region:             &region,
    }

    serviceDetails := client.ServicePOST_ServiceDetails{}
    serviceDetails.FromWebServiceDetailsPOST(webServiceDetails)

    request := client.CreateServiceJSONRequestBody{
        Name:           "my-api-service",
        OwnerId:        "own-123abc",
        Type:           client.WebService,
        ServiceDetails: &serviceDetails,
        Repo:           &repo,
        Branch:         &branch,
        EnvVars:        &envVars,
    }

    response, err := serviceRepo.CreateService(ctx, request)
    if err != nil {
        panic(err)
    }
    // Response: {"service":{"id":"srv-xyz789","name":"my-api-service","createdAt":"2025-11-01T10:00:00Z"}}
}

// List all services in workspace
func listServices(ctx context.Context, serviceRepo *service.Repo) {
    params := &client.ListServicesParams{}
    includePreviews := false
    params.IncludePreviews = &includePreviews

    services, err := serviceRepo.ListServices(ctx, params)
    if err != nil {
        panic(err)
    }
    // Response: [{"id":"srv-xyz789","name":"my-api-service","type":"web_service"}]
}

// Get specific service details
func getService(ctx context.Context, serviceRepo *service.Repo, serviceId string) {
    service, err := serviceRepo.GetService(ctx, serviceId)
    if err != nil {
        panic(err)
    }
    // Response includes full service configuration, status, and URLs
}
```

### Static Site Management

Deploy static sites with CDN distribution for frontend applications.

```go
// Create a React static site
func createStaticSite(ctx context.Context, serviceRepo *service.Repo) {
    repo := "https://github.com/username/my-react-app.git"
    branch := "main"
    buildCommand := "yarn install && yarn build"
    publishPath := "build"

    staticDetails := client.StaticSiteDetailsPOST{
        BuildCommand: &buildCommand,
        PublishPath:  &publishPath,
    }

    serviceDetails := client.ServicePOST_ServiceDetails{}
    serviceDetails.FromStaticSiteDetailsPOST(staticDetails)

    envVars := []client.EnvVarInput{}
    envVar := client.EnvVarInput{}
    envVar.FromEnvVarKeyValue(client.EnvVarKeyValue{
        Key:   "REACT_APP_API_URL",
        Value: "https://my-api-service.onrender.com",
    })
    envVars = append(envVars, envVar)

    request := client.CreateServiceJSONRequestBody{
        Name:           "my-react-frontend",
        OwnerId:        "own-123abc",
        Type:           client.StaticSite,
        ServiceDetails: &serviceDetails,
        Repo:           &repo,
        Branch:         &branch,
        EnvVars:        &envVars,
    }

    response, err := serviceRepo.CreateService(ctx, request)
    if err != nil {
        panic(err)
    }
    // Response: {"service":{"id":"srv-abc123","url":"https://my-react-frontend.onrender.com"}}
}
```

### Environment Variable Management

Update service configuration without redeployment triggers, with automatic merge support.

```go
// Update environment variables with merge strategy
func updateEnvVars(ctx context.Context, serviceRepo *service.Repo, serviceId string) {
    // New variables to set
    newVars := []client.EnvVarInput{}

    var1 := client.EnvVarInput{}
    var1.FromEnvVarKeyValue(client.EnvVarKeyValue{
        Key:   "DATABASE_URL",
        Value: "postgresql://user:pass@host:5432/db",
    })
    newVars = append(newVars, var1)

    var2 := client.EnvVarInput{}
    var2.FromEnvVarKeyValue(client.EnvVarKeyValue{
        Key:   "API_KEY",
        Value: "sk_live_abc123",
    })
    newVars = append(newVars, var2)

    // Merge with existing variables (replace=false by default)
    oldVars, err := serviceRepo.ListEnvVars(ctx, serviceId, &client.GetEnvVarsForServiceParams{})
    if err != nil {
        panic(err)
    }

    // Merge logic
    envVarMap := make(map[string]string)
    for _, envVar := range oldVars {
        envVarMap[envVar.Key] = envVar.Value
    }

    for _, envVarInput := range newVars {
        envVar, _ := envVarInput.AsEnvVarKeyValue()
        envVarMap[envVar.Key] = envVar.Value
    }

    // Update and trigger deploy
    mergedVars := []client.EnvVarInput{}
    for k, v := range envVarMap {
        var envVarInput client.EnvVarInput
        envVarInput.FromEnvVarKeyValue(client.EnvVarKeyValue{Key: k, Value: v})
        mergedVars = append(mergedVars, envVarInput)
    }

    _, err = serviceRepo.UpdateEnvVars(ctx, serviceId, mergedVars)
    if err != nil {
        panic(err)
    }

    // Automatically triggers deployment
    deployResponse, err := serviceRepo.DeployService(ctx, serviceId)
    if err != nil {
        panic(err)
    }
    // Response: {"deploy":{"id":"dep-xyz123","status":"pending"}}
}
```

### PostgreSQL Database Management

Create and query PostgreSQL databases with read-only transaction safety.

```go
// Create a PostgreSQL database
import (
    "github.com/render-oss/render-mcp-server/pkg/postgres"
)

func createPostgres(ctx context.Context, c *client.ClientWithResponses) {
    postgresRepo := postgres.NewRepo(c)

    region := "oregon"
    diskSize := 10
    version := client.PostgresVersion("16")

    createParams := client.PostgresPOSTInput{
        Name:       "my-production-db",
        OwnerId:    "own-123abc",
        Plan:       "pro_4gb",
        Region:     &region,
        Version:    version,
        DiskSizeGB: &diskSize,
    }

    pg, err := postgresRepo.CreatePostgres(ctx, createParams)
    if err != nil {
        panic(err)
    }
    // Response: {"id":"pgdb-xyz789","name":"my-production-db","status":"creating"}
}

// Query database with read-only transactions
func queryPostgres(ctx context.Context, postgresRepo *postgres.Repo, postgresId string) {
    sqlQuery := "SELECT id, email, created_at FROM users WHERE active = true ORDER BY created_at DESC LIMIT 10"

    connInfo, err := postgresRepo.GetPostgresConnectionInfo(ctx, postgresId)
    if err != nil {
        panic(err)
    }

    config, _ := pgx.ParseConfig(connInfo.ExternalConnectionString)
    conn, _ := pgx.ConnectConfig(ctx, config)
    defer conn.Close(ctx)

    // Automatically wrapped in READ ONLY transaction
    tx, _ := conn.BeginTx(ctx, pgx.TxOptions{AccessMode: pgx.ReadOnly})
    defer tx.Rollback(ctx)

    rows, err := tx.Query(ctx, sqlQuery)
    if err != nil {
        panic(err)
    }
    defer rows.Close()

    results := []map[string]interface{}{}
    fieldDescs := rows.FieldDescriptions()
    columnNames := make([]string, len(fieldDescs))
    for i, fd := range fieldDescs {
        columnNames[i] = string(fd.Name)
    }

    for rows.Next() {
        values, _ := rows.Values()
        rowMap := make(map[string]interface{})
        for i, col := range columnNames {
            rowMap[col] = values[i]
        }
        results = append(results, rowMap)
    }
    // Response: [{"id":1,"email":"user@example.com","created_at":"2025-01-15T10:00:00Z"}]
}
```

### Redis Key-Value Store Management

Deploy Redis instances with configurable eviction policies.

```go
// Create a Redis key-value store
import (
    "github.com/render-oss/render-mcp-server/pkg/keyvalue"
)

func createKeyValue(ctx context.Context, c *client.ClientWithResponses) {
    keyValueRepo := keyvalue.NewRepo(c)

    region := "oregon"
    policy := client.AllkeysLru

    createParams := client.KeyValuePOSTInput{
        Name:             "session-cache",
        OwnerId:          "own-123abc",
        Plan:             client.KeyValuePlanPro,
        Region:           &region,
        MaxmemoryPolicy:  &policy,
    }

    kv, err := keyValueRepo.CreateKeyValue(ctx, createParams)
    if err != nil {
        panic(err)
    }
    // Response: {"id":"kv-abc123","name":"session-cache","maxmemoryPolicy":"allkeys-lru"}
}

// List all key-value stores
func listKeyValue(ctx context.Context, keyValueRepo *keyvalue.Repo) {
    kvs, err := keyValueRepo.ListKeyValue(ctx, &client.ListKeyValueParams{})
    if err != nil {
        panic(err)
    }
    // Response: [{"id":"kv-abc123","name":"session-cache","plan":"pro","region":"oregon"}]
}
```

### Log Querying and Analysis

Search and filter application logs with regex support and pagination.

```go
// Query logs with filters
import (
    "time"
    "github.com/render-oss/render-mcp-server/pkg/logs"
    logsclient "github.com/render-oss/render-mcp-server/pkg/client/logs"
)

func listLogs(ctx context.Context, c *client.ClientWithResponses, ownerId string) {
    logRepo := logs.NewLogRepo(c)

    resources := []string{"srv-xyz789"}
    statusCodes := []string{"500", "502", "503"}
    textFilters := []string{"error", "exception"}
    limit := 50
    direction := logsclient.Backward

    startTime := time.Now().Add(-1 * time.Hour)
    endTime := time.Now()
    startTimeParam := client.StartTimeParam(startTime)
    endTimeParam := client.EndTimeParam(endTime)

    params := &client.ListLogsParams{
        OwnerId:    ownerId,
        Resource:   resources,
        StatusCode: &statusCodes,
        Text:       &textFilters,
        StartTime:  &startTimeParam,
        EndTime:    &endTimeParam,
        Direction:  &direction,
        Limit:      &limit,
    }

    response, err := logRepo.ListLogs(ctx, params)
    if err != nil {
        panic(err)
    }
    // Response: {"logs":[{"timestamp":"2025-11-01T10:30:00Z","text":"Error: Connection timeout","level":"error"}],"hasMore":true,"nextStartTime":"2025-11-01T10:25:00Z"}
}

// Discover available log labels
func listLogLabelValues(ctx context.Context, logRepo *logs.LogRepo, ownerId string) {
    resources := []string{"srv-xyz789"}

    params := &client.ListLogsValuesParams{
        OwnerId:  ownerId,
        Label:    client.ListLogsValuesParamsLabel("statusCode"),
        Resource: resources,
    }

    values, err := logRepo.ListLogLabelValues(ctx, params)
    if err != nil {
        panic(err)
    }
    // Response: ["200","404","500","502"]
}
```

### Performance Metrics Retrieval

Monitor resource utilization, HTTP performance, and database connections with time-series data.

```go
// Get comprehensive metrics
import (
    "github.com/render-oss/render-mcp-server/pkg/metrics"
    metricstypes "github.com/render-oss/render-mcp-server/pkg/client/metrics"
)

func getMetrics(ctx context.Context, c *client.ClientWithResponses, resourceId string) {
    metricsRepo := metrics.NewRepo(c)

    startTime := time.Now().Add(-24 * time.Hour)
    endTime := time.Now()
    startTimeParam := client.StartTimeParam(startTime)
    endTimeParam := client.EndTimeParam(endTime)
    resolution := float32(300) // 5-minute intervals
    quantile := metricstypes.Quantile(0.95)
    aggMethod := metricstypes.AVG

    metricsRequest := metrics.MetricsRequest{
        ResourceID:  resourceId,
        MetricTypes: []metrics.MetricType{
            metrics.MetricTypeCPUUsage,
            metrics.MetricTypeMemoryUsage,
            metrics.MetricTypeHTTPRequestCount,
            metrics.MetricTypeHTTPLatency,
            metrics.MetricTypeBandwidthUsage,
        },
        StartTime:                  &startTimeParam,
        EndTime:                    &endTimeParam,
        Resolution:                 &resolution,
        CpuUsageAggregationMethod:  &aggMethod,
        HttpLatencyQuantile:        &quantile,
    }

    response, err := metricsRepo.GetMetrics(ctx, metricsRequest)
    if err != nil {
        panic(err)
    }
    // Response: {"cpuUsage":{"datapoints":[{"timestamp":"2025-11-01T10:00:00Z","value":45.2}]},"httpLatency":{"datapoints":[{"timestamp":"2025-11-01T10:00:00Z","value":123.5}]}}
}

// Get metrics with HTTP filtering
func getHTTPMetrics(ctx context.Context, metricsRepo *metrics.Repo, serviceId string) {
    startTime := time.Now().Add(-1 * time.Hour)
    endTime := time.Now()
    startTimeParam := client.StartTimeParam(startTime)
    endTimeParam := client.EndTimeParam(endTime)

    host := metricstypes.HostQueryParam("api.example.com")
    path := metricstypes.PathQueryParam("/api/users")
    aggregateBy := metricstypes.HttpAggregateByStatusCode

    metricsRequest := metrics.MetricsRequest{
        ResourceID:  serviceId,
        MetricTypes: []metrics.MetricType{metrics.MetricTypeHTTPRequestCount},
        StartTime:   &startTimeParam,
        EndTime:     &endTimeParam,
        HttpHost:    &host,
        HttpPath:    &path,
        AggregateHttpRequestCountBy: &aggregateBy,
    }

    response, err := metricsRepo.GetMetrics(ctx, metricsRequest)
    if err != nil {
        panic(err)
    }
    // Response: {"httpRequestCount":{"datapoints":[{"timestamp":"2025-11-01T10:00:00Z","value":150,"statusCode":"200"}]}}
}
```

### Deployment Management

Track deployment history and status for continuous delivery workflows.

```go
// List deployment history
import (
    "github.com/render-oss/render-mcp-server/pkg/deploy"
)

func listDeploys(ctx context.Context, c *client.ClientWithResponses, serviceId string) {
    deployRepo := deploy.NewRepo(c)

    limit := 10
    cursor := ""
    params := &client.ListDeploysParams{
        Limit:  &limit,
        Cursor: &cursor,
    }

    deploys, nextCursor, err := deployRepo.ListDeploys(ctx, serviceId, params)
    if err != nil {
        panic(err)
    }
    // Response: [{"id":"dep-123","status":"live","createdAt":"2025-11-01T09:00:00Z","commit":"abc123"}]
    // nextCursor: "dep-122" for pagination
}

// Get specific deployment details
func getDeploy(ctx context.Context, deployRepo *deploy.Repo, serviceId string, deployId string) {
    deployment, err := deployRepo.GetDeploy(ctx, serviceId, deployId)
    if err != nil {
        panic(err)
    }
    // Response: {"id":"dep-123","status":"live","createdAt":"2025-11-01T09:00:00Z","finishedAt":"2025-11-01T09:05:23Z"}
}
```

### MCP Server Initialization

Bootstrap the MCP server with all tools and configure transport mechanisms.

```go
// Start MCP server with stdio transport
import (
    "github.com/mark3labs/mcp-go/server"
    "github.com/render-oss/render-mcp-server/cmd"
    "github.com/render-oss/render-mcp-server/pkg/cfg"
)

func main() {
    // Initialize MCP server
    s := server.NewMCPServer("render-mcp-server", cfg.Version)

    // Create authenticated client
    c, err := client.NewDefaultClient()
    if err != nil {
        panic(err)
    }

    // Register all tool categories
    s.AddTools(owner.Tools(c)...)
    s.AddTools(service.Tools(c)...)
    s.AddTools(deploy.Tools(c)...)
    s.AddTools(postgres.Tools(c)...)
    s.AddTools(keyvalue.Tools(c)...)
    s.AddTools(logs.Tools(c)...)
    s.AddTools(metrics.Tools(c)...)

    // Start with stdio transport for CLI integration
    err = server.ServeStdio(s, server.WithStdioContextFunc(
        multicontext.MultiStdioContextFunc(
            session.ContextWithStdioSession,
            authn.ContextWithAPITokenFromConfig,
        ),
    ))
    if err != nil {
        panic(err)
    }
}

// Start with HTTP transport for web integration
func startHTTPServer(s *server.MCPServer) {
    sessionStore := session.NewInMemoryStore()

    // Use Redis for distributed sessions
    if redisURL := os.Getenv("REDIS_URL"); redisURL != "" {
        sessionStore, _ = session.NewRedisStore(redisURL)
    }

    err := server.NewStreamableHTTPServer(s,
        server.WithHTTPContextFunc(multicontext.MultiHTTPContextFunc(
            session.ContextWithHTTPSession(sessionStore),
            authn.ContextWithAPITokenFromHeader,
        )),
    ).Start(":10000")
    if err != nil {
        panic(err)
    }
}
```

## Integration and Use Cases

The Render MCP Server enables AI-assisted infrastructure management through natural language interactions. Primary use cases include automated service provisioning where developers describe requirements and the LLM creates appropriate web services, databases, and caches with proper configuration; production troubleshooting where AI analyzes logs and metrics to identify error patterns, performance bottlenecks, and resource constraints; database operations allowing direct SQL queries for data analysis within conversational workflows; and deployment monitoring to track release history and status.

Integration patterns support desktop AI assistants using stdio transport for interactive terminal-based workflows, web applications using HTTP transport with session management for multi-user environments, and CI/CD pipelines for automated infrastructure operations. The server implements comprehensive error handling, read-only transaction safety for database queries, automatic workspace selection, environment variable merging to prevent accidental overwrites, and paginated responses for large datasets. Authentication uses Render API tokens configured through environment variables or HTTP headers, with context propagation ensuring workspace isolation across all operations.