@aibuilders/mcp-coach-server 1.0.5 → 1.0.7

package/dist/index.js

@@ -4,9 +4,16 @@ import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"
 import { CallToolRequestSchema, ListToolsRequestSchema, } from "@modelcontextprotocol/sdk/types.js";
 import dotenv from "dotenv";
 import { readFile, writeFile, access, mkdir } from "fs/promises";
-import { join } from "path";
+import { readFileSync } from "fs";
+import { join, dirname } from "path";
 import { homedir } from "os";
+import { fileURLToPath } from "url";
 dotenv.config();
+// Get package version from package.json
+const __filename = fileURLToPath(import.meta.url);
+const __dirname = dirname(__filename);
+const packageJson = JSON.parse(readFileSync(join(__dirname, "../package.json"), "utf-8"));
+const PACKAGE_VERSION = packageJson.version;
 const CACHE_DIR = join(homedir(), ".ai-builders-mcp-cache");
 const DEPLOYMENT_GUIDE_CACHE = join(CACHE_DIR, "deployment_guide_cache.json");
 async function ensureCacheDir() {
@@ -93,7 +100,7 @@ async function getCachedDeploymentGuide() {
 }
 const server = new Server({
     name: "ai-builder-mcp",
-    version: "1.0.5",
+    version: PACKAGE_VERSION,
 }, {
     capabilities: {
         tools: {},
@@ -116,10 +123,8 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
                         if (url.startsWith("http")) {
                             baseUrl = url;
                         }
-                        else if (url.startsWith("/")) {
-                            // Relative path: use the backend base URL directly
-                            // Don't append the servers path, use the base URL as-is
-                            baseUrl = "https://space.ai-builders.com/backend";
+                        else {
+                            baseUrl = `https://space.ai-builders.com${url}`;
                         }
                     }
                 }
@@ -192,9 +197,42 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
                                             "1. Check if a deploy-config.json file exists in the project",
                                             "2. If not, help create one with deployment parameters (repo_url, service_name, branch, port) and environment variables",
                                             "3. Remind users not to commit sensitive values (passwords, API keys) to Git repositories",
-                                            "4. When user says 'deploy', read from deploy-config.json and include env_vars in the API request",
-                                            "5. Guide users to maintain their configuration for future deployments"
+                                            "4. Before deploying, help users test the Dockerfile locally:",
+                                            "   - Build the Docker image: docker build -t test-image .",
+                                            "   - Run the container locally: docker run -p <port>:<port> -e PORT=<port> test-image",
+                                            "   - Verify the service starts correctly and responds to requests",
+                                            "   - Test with the same environment variables that will be used in production",
+                                            "5. When user says 'deploy', read from deploy-config.json and include env_vars in the API request",
+                                            "6. After deployment, monitor the deployment status. If deployment fails:",
+                                            "   - Use GET /v1/deployments/{service_name}/logs to retrieve deployment logs",
+                                            "   - Check build logs: GET /v1/deployments/{service_name}/logs?log_type=build",
+                                            "   - Check runtime logs: GET /v1/deployments/{service_name}/logs?log_type=runtime",
+                                            "   - Filter by stream: GET /v1/deployments/{service_name}/logs?stream=stderr (for errors)",
+                                            "   - Help users debug based on log output and fix issues",
+                                            "7. Guide users to maintain their configuration for future deployments"
                                         ],
+                                        log_api_info: {
+                                            endpoint: "GET /v1/deployments/{service_name}/logs",
+                                            description: "Retrieve deployment logs from Koyeb for debugging failed deployments",
+                                            parameters: {
+                                                service_name: "Path parameter: The unique service name / subdomain",
+                                                log_type: "Query parameter: 'build' or 'runtime' (default: 'runtime')",
+                                                stream: "Query parameter: Filter by 'stdout', 'stderr', or 'koyeb'",
+                                                timeout: "Query parameter: Seconds to wait for streaming logs (1-300, default: 5)",
+                                                deployment_id: "Query parameter: Optional deployment ID override"
+                                            },
+                                            usage_examples: {
+                                                build_logs: "GET /v1/deployments/my-service/logs?log_type=build",
+                                                runtime_errors: "GET /v1/deployments/my-service/logs?log_type=runtime&stream=stderr",
+                                                all_logs: "GET /v1/deployments/my-service/logs?log_type=runtime&timeout=30"
+                                            },
+                                            when_to_use: [
+                                                "When deployment status shows ERROR, UNHEALTHY, or DEGRADED",
+                                                "To debug why a service is not starting correctly",
+                                                "To check build-time errors during Docker image creation",
+                                                "To monitor runtime errors and application logs"
+                                            ]
+                                        },
                                         example_deploy_config: {
                                             description: "Example deploy-config.json structure",
                                             content: {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@aibuilders/mcp-coach-server",
-  "version": "1.0.5",
+  "version": "1.0.7",
   "description": "AI Builder MCP server for AI Builders platform",
   "main": "dist/index.js",
   "type": "module",