Add ScrapeGraphAI registry support with PyPI and Smithery

docs/modelcontextprotocol-io/package-types.mdx

@@ -162,6 +162,86 @@ The MCP Registry verifies ownership of Docker/OCI images by checking for an `io.
 LABEL io.modelcontextprotocol.server.name="io.github.username/kubernetes-manager-mcp"
 ```
 
+## ScrapeGraphAI Packages
+
+For ScrapeGraphAI packages, the MCP Registry currently supports packages distributed via:
+- The official PyPI registry (`https://pypi.org`)
+- Smithery (`https://smithery.ai`), which distributes packages via npm
+
+ScrapeGraphAI packages use `"registryType": "scrapegraphai"` in `server.json`. For example:
+
+**Example - ScrapeGraphAI package from PyPI:**
+
+```json server.json highlight={9}
+{
+  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
+  "name": "io.github.username/web-scraper-mcp",
+  "title": "Web Scraper",
+  "description": "Scrape and extract data from websites using ScrapeGraphAI",
+  "version": "1.0.0",
+  "packages": [
+    {
+      "registryType": "scrapegraphai",
+      "registryBaseUrl": "https://pypi.org",
+      "identifier": "web-scraper-mcp",
+      "version": "1.0.0",
+      "transport": {
+        "type": "stdio"
+      }
+    }
+  ]
+}
+```
+
+**Example - ScrapeGraphAI package from Smithery:**
+
+```json server.json highlight={9}
+{
+  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
+  "name": "io.scrapegraphai/scrapegraph-mcp",
+  "title": "ScrapeGraph MCP",
+  "description": "Scrape and extract data from websites using ScrapeGraphAI",
+  "version": "1.0.0",
+  "packages": [
+    {
+      "registryType": "scrapegraphai",
+      "registryBaseUrl": "https://smithery.ai",
+      "identifier": "@ScrapeGraphAI/scrapegraph-mcp",
+      "version": "1.0.0",
+      "transport": {
+        "type": "stdio"
+      }
+    }
+  ]
+}
+```
+
+### Ownership Verification
+
+**For PyPI packages:** The MCP Registry verifies ownership by checking for the existence of an `mcp-name: $SERVER_NAME` string in the package README (which becomes the package description on PyPI). The string may be hidden in a comment, but the `$SERVER_NAME` portion **MUST** match the server name from `server.json`.
+
+**For Smithery packages:** The MCP Registry verifies ownership by checking the `mcpName` field in `package.json` (since Smithery uses npm as backend). The `mcpName` property **MUST** match the server name from `server.json`.
+
+**Example for PyPI:**
+
+```markdown README.md highlight={5}
+# Web Scraper MCP Server
+
+This MCP server scrapes and extracts data from websites using ScrapeGraphAI.
+
+<!-- mcp-name: io.github.username/web-scraper-mcp -->
+```
+
+**Example for Smithery (npm):**
+
+```json package.json
+{
+  "name": "@ScrapeGraphAI/scrapegraph-mcp",
+  "version": "1.0.0",
+  "mcpName": "io.scrapegraphai/scrapegraph-mcp"
+}
+```
+
 ## MCPB Packages
 
 For MCPB packages, the MCP Registry currently supports MCPB artifacts hosted via GitHub or GitLab releases.

docs/reference/api/openapi.yaml

@@ -488,13 +488,14 @@ components:
       properties:
         registryType:
           type: string
-          description: Registry type indicating how to download packages (e.g., 'npm', 'pypi', 'oci', 'nuget', 'mcpb')
+          description: Registry type indicating how to download packages (e.g., 'npm', 'pypi', 'oci', 'nuget', 'mcpb', 'scrapegraphai')
           examples:
             - "npm"
             - "pypi"
             - "oci"
             - "nuget"
             - "mcpb"
+            - "scrapegraphai"
         registryBaseUrl:
           type: string
           format: uri
@@ -506,6 +507,7 @@ components:
             - "https://api.nuget.org/v3/index.json"
             - "https://github.com"
             - "https://gitlab.com"
+            - "https://smithery.ai"
         identifier:
           type: string
           description: Package identifier - either a package name (for registries) or URL (for direct downloads)

docs/reference/server-json/server.schema.json

@@ -242,19 +242,21 @@
             "https://docker.io",
             "https://api.nuget.org/v3/index.json",
             "https://github.com",
-            "https://gitlab.com"
+            "https://gitlab.com",
+            "https://smithery.ai"
           ],
           "format": "uri",
           "type": "string"
         },
         "registryType": {
-          "description": "Registry type indicating how to download packages (e.g., 'npm', 'pypi', 'oci', 'nuget', 'mcpb')",
+          "description": "Registry type indicating how to download packages (e.g., 'npm', 'pypi', 'oci', 'nuget', 'mcpb', 'scrapegraphai')",
           "examples": [
             "npm",
             "pypi",
             "oci",
             "nuget",
-            "mcpb"
+            "mcpb",
+            "scrapegraphai"
           ],
           "type": "string"
         },

internal/validators/package.go

@@ -23,6 +23,8 @@ func ValidatePackage(ctx context.Context, pkg model.Package, serverName string)
 		return registries.ValidateOCI(ctx, pkg, serverName)
 	case model.RegistryTypeMCPB:
 		return registries.ValidateMCPB(ctx, pkg, serverName)
+	case model.RegistryTypeScrapeGraphAI:
+		return registries.ValidateScrapeGraphAI(ctx, pkg, serverName)
 	default:
 		return fmt.Errorf("unsupported registry type: %s", pkg.RegistryType)
 	}

internal/validators/registries/scrapegraphai.go

@@ -0,0 +1,142 @@
+package registries
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+	"net/url"
+	"strings"
+	"time"
+
+	"github.com/modelcontextprotocol/registry/pkg/model"
+)
+
+var (
+	ErrMissingIdentifierForScrapeGraphAI = errors.New("package identifier is required for ScrapeGraphAI packages")
+	ErrMissingVersionForScrapeGraphAI   = errors.New("package version is required for ScrapeGraphAI packages")
+)
+
+// ScrapeGraphAIPackageResponse represents the structure returned by the PyPI JSON API
+// (ScrapeGraphAI packages are distributed via PyPI)
+type ScrapeGraphAIPackageResponse struct {
+	Info struct {
+		Description string `json:"description"`
+	} `json:"info"`
+}
+
+// ValidateScrapeGraphAI validates that a ScrapeGraphAI package contains the correct MCP server name
+// ScrapeGraphAI packages can be distributed via PyPI or via Smithery (which uses npm)
+func ValidateScrapeGraphAI(ctx context.Context, pkg model.Package, serverName string) error {
+	// Set default registry base URL if empty (use PyPI since ScrapeGraphAI is on PyPI)
+	if pkg.RegistryBaseURL == "" {
+		pkg.RegistryBaseURL = model.RegistryURLPyPI
+	}
+
+	if pkg.Identifier == "" {
+		return ErrMissingIdentifierForScrapeGraphAI
+	}
+
+	if pkg.Version == "" {
+		return ErrMissingVersionForScrapeGraphAI
+	}
+
+	// Validate that MCPB-specific fields are not present
+	if pkg.FileSHA256 != "" {
+		return fmt.Errorf("ScrapeGraphAI packages must not have 'fileSha256' field - this is only for MCPB packages")
+	}
+
+	// Validate registry base URL - support both PyPI and Smithery (which uses npm)
+	if pkg.RegistryBaseURL != model.RegistryURLPyPI && pkg.RegistryBaseURL != model.RegistryURLSmithery && pkg.RegistryBaseURL != model.RegistryURLNPM {
+		return fmt.Errorf("registry type and base URL do not match: '%s' is not valid for registry type '%s'. Supported URLs: %s, %s, %s",
+			pkg.RegistryBaseURL, model.RegistryTypeScrapeGraphAI, model.RegistryURLPyPI, model.RegistryURLSmithery, model.RegistryURLNPM)
+	}
+
+	client := &http.Client{Timeout: 10 * time.Second}
+
+	// If using Smithery or npm, validate via npm API (Smithery uses npm as backend)
+	if pkg.RegistryBaseURL == model.RegistryURLSmithery || pkg.RegistryBaseURL == model.RegistryURLNPM {
+		return validateScrapeGraphAIViaNPM(ctx, client, pkg, serverName)
+	}
+
+	// Otherwise, validate via PyPI API
+	return validateScrapeGraphAIViaPyPI(ctx, client, pkg, serverName)
+}
+
+// validateScrapeGraphAIViaPyPI validates ScrapeGraphAI packages distributed via PyPI
+func validateScrapeGraphAIViaPyPI(ctx context.Context, client *http.Client, pkg model.Package, serverName string) error {
+	requestURL := fmt.Sprintf("%s/pypi/%s/%s/json", pkg.RegistryBaseURL, pkg.Identifier, pkg.Version)
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, requestURL, nil)
+	if err != nil {
+		return fmt.Errorf("failed to create request: %w", err)
+	}
+
+	req.Header.Set("User-Agent", "MCP-Registry-Validator/1.0")
+	req.Header.Set("Accept", "application/json")
+
+	resp, err := client.Do(req)
+	if err != nil {
+		return fmt.Errorf("failed to fetch package metadata from PyPI for ScrapeGraphAI package: %w", err)
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != http.StatusOK {
+		return fmt.Errorf("ScrapeGraphAI package '%s' not found on PyPI (status: %d)", pkg.Identifier, resp.StatusCode)
+	}
+
+	var scrapeGraphAIResp ScrapeGraphAIPackageResponse
+	if err := json.NewDecoder(resp.Body).Decode(&scrapeGraphAIResp); err != nil {
+		return fmt.Errorf("failed to parse ScrapeGraphAI package metadata: %w", err)
+	}
+
+	// Check description (README) content
+	description := scrapeGraphAIResp.Info.Description
+
+	// Check for mcp-name: format (more specific)
+	mcpNamePattern := "mcp-name: " + serverName
+	if strings.Contains(description, mcpNamePattern) {
+		return nil // Found as mcp-name: format
+	}
+
+	return fmt.Errorf("ScrapeGraphAI package '%s' ownership validation failed. The server name '%s' must appear as 'mcp-name: %s' in the package README", pkg.Identifier, serverName, serverName)
+}
+
+// validateScrapeGraphAIViaNPM validates ScrapeGraphAI packages distributed via npm (Smithery uses npm)
+func validateScrapeGraphAIViaNPM(ctx context.Context, client *http.Client, pkg model.Package, serverName string) error {
+	// Use npm registry for validation (Smithery packages are on npm)
+	// Always use npm registry URL since Smithery uses npm as backend
+	requestURL := model.RegistryURLNPM + "/" + url.PathEscape(pkg.Identifier) + "/" + url.PathEscape(pkg.Version)
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, requestURL, nil)
+	if err != nil {
+		return fmt.Errorf("failed to create request: %w", err)
+	}
+
+	req.Header.Set("User-Agent", "MCP-Registry-Validator/1.0")
+	req.Header.Set("Accept", "application/json")
+
+	resp, err := client.Do(req)
+	if err != nil {
+		return fmt.Errorf("failed to fetch package metadata from npm for ScrapeGraphAI package: %w", err)
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != http.StatusOK {
+		return fmt.Errorf("ScrapeGraphAI package '%s' not found on npm/Smithery (status: %d)", pkg.Identifier, resp.StatusCode)
+	}
+
+	var npmResp NPMPackageResponse
+	if err := json.NewDecoder(resp.Body).Decode(&npmResp); err != nil {
+		return fmt.Errorf("failed to parse ScrapeGraphAI package metadata from npm: %w", err)
+	}
+
+	if npmResp.MCPName == "" {
+		return fmt.Errorf("ScrapeGraphAI package '%s' is missing required 'mcpName' field. Add this to your package.json: \"mcpName\": \"%s\"", pkg.Identifier, serverName)
+	}
+
+	if npmResp.MCPName != serverName {
+		return fmt.Errorf("ScrapeGraphAI package ownership validation failed. Expected mcpName '%s', got '%s'", serverName, npmResp.MCPName)
+	}
+
+	return nil
+}

internal/validators/registries/scrapegraphai_test.go

@@ -0,0 +1,100 @@
+package registries_test
+
+import (
+	"context"
+	"testing"
+
+	"github.com/modelcontextprotocol/registry/internal/validators/registries"
+	"github.com/modelcontextprotocol/registry/pkg/model"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestValidateScrapeGraphAI_RealPackages(t *testing.T) {
+	ctx := context.Background()
+
+	tests := []struct {
+		name         string
+		packageName  string
+		version      string
+		registryURL  string
+		serverName   string
+		expectError  bool
+		errorMessage string
+	}{
+		{
+			name:         "empty package identifier should fail",
+			packageName:  "",
+			version:      "1.0.0",
+			registryURL:  model.RegistryURLPyPI,
+			serverName:   "com.example/test",
+			expectError:  true,
+			errorMessage: "package identifier is required for ScrapeGraphAI packages",
+		},
+		{
+			name:         "empty package version should fail",
+			packageName:  "scrapegraph-mcp",
+			version:      "",
+			registryURL:  model.RegistryURLPyPI,
+			serverName:   "com.example/test",
+			expectError:  true,
+			errorMessage: "package version is required for ScrapeGraphAI packages",
+		},
+		{
+			name:         "non-existent package on PyPI should fail",
+			packageName:  generateRandomPackageName(),
+			version:      "1.0.0",
+			registryURL:  model.RegistryURLPyPI,
+			serverName:   "com.example/test",
+			expectError:  true,
+			errorMessage: "not found",
+		},
+		{
+			name:         "real PyPI package without MCP server name should fail",
+			packageName:  "requests",
+			version:      "2.31.0",
+			registryURL:  model.RegistryURLPyPI,
+			serverName:   "com.example/test",
+			expectError:  true,
+			errorMessage: "ownership validation failed",
+		},
+		{
+			name:        "real PyPI package with server name in README should pass",
+			packageName: "time-mcp-pypi",
+			version:     "1.0.6",
+			registryURL: model.RegistryURLPyPI,
+			serverName:  "io.github.domdomegg/time-mcp-pypi",
+			expectError: false,
+		},
+		{
+			name:         "invalid registry URL should fail",
+			packageName:  "scrapegraph-mcp",
+			version:      "1.0.0",
+			registryURL:  "https://invalid-registry.com",
+			serverName:   "com.example/test",
+			expectError:  true,
+			errorMessage: "registry type and base URL do not match",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			pkg := model.Package{
+				RegistryType:    model.RegistryTypeScrapeGraphAI,
+				RegistryBaseURL: tt.registryURL,
+				Identifier:      tt.packageName,
+				Version:         tt.version,
+			}
+
+			err := registries.ValidateScrapeGraphAI(ctx, pkg, tt.serverName)
+
+			if tt.expectError {
+				assert.Error(t, err)
+				if tt.errorMessage != "" {
+					assert.Contains(t, err.Error(), tt.errorMessage)
+				}
+			} else {
+				assert.NoError(t, err)
+			}
+		})
+	}
+}

internal/validators/validators_test.go

@@ -1767,6 +1767,10 @@ func TestValidate_RegistryTypesAndUrls(t *testing.T) {
 		{"valid_nuget", "io.github.domdomegg/time-mcp-server", model.RegistryTypeNuGet, "", "TimeMcpServer", "1.0.2", "", false},
 		{"valid_mcpb_github", "io.github.domdomegg/airtable-mcp-server", model.RegistryTypeMCPB, "", "https://github.com/domdomegg/airtable-mcp-server/releases/download/v1.7.2/airtable-mcp-server.mcpb", "", "fe333e598595000ae021bd27117db32ec69af6987f507ba7a63c90638ff633ce", false},
 		{"valid_mcpb_gitlab", "io.gitlab.fforster/gitlab-mcp", model.RegistryTypeMCPB, "", "https://gitlab.com/fforster/gitlab-mcp/-/releases/v1.31.0/downloads/gitlab-mcp_1.31.0_Linux_x86_64.tar.gz", "", "abc123ef4567890abcdef1234567890abcdef1234567890abcdef1234567890", false}, // this is not actually a valid mcpb, but it's the closest I can get for testing for now
+		{"valid_scrapegraphai_pypi", "io.github.domdomegg/time-mcp-pypi", model.RegistryTypeScrapeGraphAI, model.RegistryURLPyPI, "time-mcp-pypi", "1.0.6", "", false},
+		{"valid_scrapegraphai_pypi_default", "io.github.domdomegg/time-mcp-pypi", model.RegistryTypeScrapeGraphAI, "", "time-mcp-pypi", "1.0.6", "", false},
+		{"valid_scrapegraphai_smithery", "io.scrapegraphai/scrapegraph-mcp", model.RegistryTypeScrapeGraphAI, model.RegistryURLSmithery, "@ScrapeGraphAI/scrapegraph-mcp", "1.0.0", "", false},
+		{"valid_scrapegraphai_npm", "io.scrapegraphai/scrapegraph-mcp", model.RegistryTypeScrapeGraphAI, model.RegistryURLNPM, "@ScrapeGraphAI/scrapegraph-mcp", "1.0.0", "", false},
 
 		// Test MCPB without file hash (should fail)
 		{"invalid_mcpb_no_hash", "io.github.domdomegg/airtable-mcp-server", model.RegistryTypeMCPB, "", "https://github.com/domdomegg/airtable-mcp-server/releases/download/v1.7.2/airtable-mcp-server.mcpb", "", "", true},