@aws/ml-container-creator 0.9.0 → 0.9.1

package/bin/cli.js

@@ -6,6 +6,7 @@ import { createRequire } from 'module';
 import path from 'path';
 import { program, Option, Help } from 'commander';
 import { run } from '../src/app.js';
+import { cliOptions, helpGroups } from '../src/lib/generated/cli-options.js';
 
 const require = createRequire(import.meta.url);
 const { version } = require('../package.json');
@@ -23,116 +24,32 @@ program
     .version(version)
     .enablePositionalOptions()
     .helpCommand('help [command]', 'Display help for command')
-    .argument('[project-name...]', 'Name for the generated project')
-
-    // --- General ---
-    .addOption(new Option('--skip-prompts', 'Skip interactive prompts and use configuration from other sources'))
-    .addOption(new Option('--auto-prompt', 'Fill defaults, prompt only for missing required values'))
-    .addOption(new Option('--config <path>', 'Path to JSON configuration file'))
-    .addOption(new Option('--project-name <name>', 'Project name'))
-    .addOption(new Option('--project-dir <dir>', 'Output directory path'))
-    .addOption(new Option('--force', 'Overwrite existing output directory without prompting'))
-
-    // --- Model & Framework ---
-    .addOption(new Option('--deployment-config <config>', 'Deployment configuration (e.g. http-flask, transformers-vllm, triton-fil)'))
-    .addOption(new Option('--framework <framework>', 'ML framework — DEPRECATED: use --deployment-config').choices(['sklearn', 'xgboost', 'tensorflow', 'transformers']).hideHelp())
-    .addOption(new Option('--model-format <format>', 'Model serialization format (pkl, joblib, json, model, ubj, keras, h5, SavedModel)'))
-    .addOption(new Option('--model-name <name>', 'Model identifier (<hf-org/model>, s3://..., registry://..., marketplace://...)'))
-    .addOption(new Option('--model-server <server>', 'Model server — DEPRECATED: use --deployment-config').choices(['flask', 'fastapi', 'vllm', 'sglang']).hideHelp())
-    .addOption(new Option('--base-image <image>', 'Base container image for Dockerfile'))
-
-    // --- Build & Infrastructure ---
-    .addOption(new Option('--deployment-target <target>', 'Deployment target (realtime-inference, async-inference, batch-transform, hyperpod-eks)'))
-    .addOption(new Option('--instance-type <type>', 'SageMaker instance type (e.g. ml.g5.xlarge, ml.m5.large)'))
-    .addOption(new Option('--region <region>', 'AWS region'))
-    .addOption(new Option('--role-arn <arn>', 'IAM role ARN for SageMaker execution'))
-    .addOption(new Option('--build-target <target>', 'Build target (codebuild)'))
-    .addOption(new Option('--codebuild-compute-type <type>', 'CodeBuild compute type (SMALL, MEDIUM, LARGE)'))
-
-    // --- Endpoint (Real-Time Inference) ---
-    .addOption(new Option('--endpoint-initial-instance-count <n>', 'Number of instances for the endpoint (default: 1)'))
-    .addOption(new Option('--endpoint-data-capture-percent <pct>', 'Data capture percentage for monitoring, 0-100 (default: 0)'))
-    .addOption(new Option('--endpoint-variant-name <name>', 'Production variant name (default: AllTraffic)'))
-    .addOption(new Option('--endpoint-volume-size <gb>', 'ML storage volume size in GB'))
-
-    // --- Inference Component ---
-    .addOption(new Option('--ic-cpu-count <n>', 'vCPUs allocated to the inference component'))
-    .addOption(new Option('--ic-memory-size <mb>', 'Memory in MB for the inference component'))
-    .addOption(new Option('--ic-gpu-count <n>', 'GPUs allocated to the inference component'))
-    .addOption(new Option('--ic-copy-count <n>', 'Number of inference component copies (default: 1)'))
-    .addOption(new Option('--ic-model-weight <weight>', 'Traffic routing weight, 0-1 (default: 1.0)'))
-
-    // --- Async Inference ---
-    .addOption(new Option('--async-s3-output-path <path>', 'S3 output path for async results'))
-    .addOption(new Option('--async-sns-success-topic <arn>', 'SNS topic ARN for success notifications'))
-    .addOption(new Option('--async-sns-error-topic <arn>', 'SNS topic ARN for error notifications'))
-    .addOption(new Option('--async-max-concurrent <n>', 'Max concurrent invocations per instance (default: 1)'))
-
-    // --- Batch Transform ---
-    .addOption(new Option('--batch-input-path <path>', 'S3 input path for batch data'))
-    .addOption(new Option('--batch-output-path <path>', 'S3 output path for batch results'))
-    .addOption(new Option('--batch-instance-count <n>', 'Number of instances (default: 1)'))
-    .addOption(new Option('--batch-split-type <type>', 'Input split type: Line, RecordIO, None (default: Line)'))
-    .addOption(new Option('--batch-strategy <strategy>', 'Batch strategy: MultiRecord, SingleRecord (default: MultiRecord)'))
-    .addOption(new Option('--batch-join-source <source>', 'Join source: Input, None (default: None)'))
-    .addOption(new Option('--batch-max-concurrent <n>', 'Max concurrent transforms per instance (default: 1)'))
-    .addOption(new Option('--batch-max-payload <mb>', 'Max payload size in MB, 0-100 (default: 6)'))
-
-    // --- HyperPod (EKS) ---
-    .addOption(new Option('--hyperpod-cluster <name>', 'HyperPod EKS cluster name'))
-    .addOption(new Option('--hyperpod-namespace <ns>', 'Kubernetes namespace (default: default)'))
-    .addOption(new Option('--hyperpod-replicas <count>', 'Number of replicas (default: 1)'))
-    .addOption(new Option('--fsx-volume-handle <handle>', 'FSx for Lustre volume handle'))
-
-    // --- Environment Variables ---
-    .addOption(new Option('--model-env <KEY=VALUE>', 'Model env var, repeatable (e.g. VLLM_TENSOR_PARALLEL_SIZE=4)').argParser(collect).default([]))
-    .addOption(new Option('--server-env <KEY=VALUE>', 'Server env var, repeatable (e.g. SGLANG_MEM_FRACTION=0.9)').argParser(collect).default([]))
-
-    // --- Authentication ---
-    .addOption(new Option('--hf-token <token>', 'HuggingFace token (or "$HF_TOKEN" for env var reference)'))
-    .addOption(new Option('--hf-token-arn <arn>', 'HuggingFace token ARN from Secrets Manager'))
-    .addOption(new Option('--ngc-token <token>', 'NVIDIA NGC token (or "$NGC_API_KEY" for env var reference)'))
-    .addOption(new Option('--ngc-token-arn <arn>', 'NVIDIA NGC token ARN from Secrets Manager'))
-
-    // --- Optional Features ---
-    .addOption(new Option('--include-sample', 'Include sample model code'))
-    .addOption(new Option('--include-testing', 'Include test suite'))
-    .addOption(new Option('--test-types <types>', 'Comma-separated test types'))
-    .addOption(new Option('--enable-lora', 'Enable LoRA adapter serving (transformers with vllm/sglang/djl-lmi only)'))
-    .addOption(new Option('--max-loras <n>', 'Maximum concurrent LoRA adapters in GPU memory (default: 30)'))
-    .addOption(new Option('--max-lora-rank <n>', 'Maximum LoRA rank (default: 64)'))
-
-    // --- Benchmarking ---
-    .addOption(new Option('--include-benchmark', 'Include SageMaker AI Benchmarking (transformers/diffusors only)'))
-    .addOption(new Option('--benchmark-concurrency <n>', 'Benchmark concurrent requests (default: 10)'))
-    .addOption(new Option('--benchmark-input-tokens <n>', 'Benchmark mean input tokens (default: 550)'))
-    .addOption(new Option('--benchmark-output-tokens <n>', 'Benchmark mean output tokens (default: 150)'))
-    .addOption(new Option('--benchmark-streaming', 'Enable streaming in benchmark (default: true)'))
-    .addOption(new Option('--benchmark-request-count <n>', 'Total benchmark requests (optional)'))
-    .addOption(new Option('--benchmark-s3-output-path <path>', 'S3 path for benchmark results'))
-
-    // --- MCP & Discovery ---
-    .addOption(new Option('--smart', 'Enable Bedrock-powered smart mode on MCP servers'))
-    .addOption(new Option('--discover', 'Enable live registry lookups via MCP discovery'))
-
-    // --- Validation ---
-    .addOption(new Option('--no-validate', 'Skip schema-driven validation at generation time'))
-    .addOption(new Option('--validate-env-vars', 'Enable environment variable validation (default: true)'))
-    .addOption(new Option('--validate-with-docker', 'Enable Docker introspection validation (opt-in)'))
-    .addOption(new Option('--offline', 'Disable HuggingFace API lookups'))
-
-    .action((projectNameArgs, options) => {
-        // Mutual exclusion validation: plaintext token and ARN flags cannot both be provided
-        if (options.hfToken && options.hfTokenArn) {
-            console.error('❌ Cannot specify both --hf-token and --hf-token-arn. Use one or the other.');
-            process.exit(1);
-        }
-        if (options.ngcToken && options.ngcTokenArn) {
-            console.error('❌ Cannot specify both --ngc-token and --ngc-token-arn. Use one or the other.');
-            process.exit(1);
-        }
-        return run(projectNameArgs?.[0] || null, options);
-    });
+    .argument('[project-name...]', 'Name for the generated project');
+
+// Register all CLI options from generated schema
+for (const opt of cliOptions) {
+    if (opt.hidden) continue;
+    const option = new Option(opt.flag, opt.description);
+    if (opt.choices) option.choices(opt.choices);
+    if (opt.repeatable) {
+        option.argParser(collect);
+        option.default([]);
+    }
+    program.addOption(option);
+}
+
+program.action((projectNameArgs, options) => {
+    // Mutual exclusion validation: plaintext token and ARN flags cannot both be provided
+    if (options.hfToken && options.hfTokenArn) {
+        console.error('❌ Cannot specify both --hf-token and --hf-token-arn. Use one or the other.');
+        process.exit(1);
+    }
+    if (options.ngcToken && options.ngcTokenArn) {
+        console.error('❌ Cannot specify both --ngc-token and --ngc-token-arn. Use one or the other.');
+        process.exit(1);
+    }
+    return run(projectNameArgs?.[0] || null, options);
+});
 
 // Custom help formatting — group options into logical sections (root command only)
 program.configureHelp({
@@ -163,7 +80,7 @@ program.configureHelp({
         // Collect all visible options
         const allOptions = helper.visibleOptions(cmd);
 
-        // Partition options into groups by flag prefix/purpose
+        // Partition options into groups using schema-derived helpGroups
         const groups = {
             general: [],
             model: [],
@@ -182,32 +99,9 @@ program.configureHelp({
 
         for (const opt of allOptions) {
             const long = opt.long || '';
-            if (['--skip-prompts', '--auto-prompt', '--config', '--project-name', '--project-dir', '--force', '--version', '--help'].includes(long)) {
-                groups.general.push(opt);
-            } else if (['--deployment-config', '--framework', '--model-format', '--model-name', '--model-server', '--base-image'].includes(long)) {
-                groups.model.push(opt);
-            } else if (['--deployment-target', '--instance-type', '--region', '--role-arn', '--build-target', '--codebuild-compute-type'].includes(long)) {
-                groups.infra.push(opt);
-            } else if (long.startsWith('--endpoint-')) {
-                groups.endpoint.push(opt);
-            } else if (long.startsWith('--ic-')) {
-                groups.ic.push(opt);
-            } else if (long.startsWith('--async-')) {
-                groups.async.push(opt);
-            } else if (long.startsWith('--batch-')) {
-                groups.batch.push(opt);
-            } else if (long.startsWith('--hyperpod-') || long === '--fsx-volume-handle') {
-                groups.hyperpod.push(opt);
-            } else if (['--model-env', '--server-env'].includes(long)) {
-                groups.env.push(opt);
-            } else if (['--hf-token', '--hf-token-arn', '--ngc-token', '--ngc-token-arn'].includes(long)) {
-                groups.auth.push(opt);
-            } else if (['--include-sample', '--include-testing', '--test-types', '--enable-lora', '--max-loras', '--max-lora-rank'].includes(long)) {
-                groups.features.push(opt);
-            } else if (['--smart', '--discover'].includes(long)) {
-                groups.mcp.push(opt);
-            } else if (['--validate-env-vars', '--validate-with-docker', '--offline', '--no-validate'].includes(long)) {
-                groups.validation.push(opt);
+            const section = helpGroups[long] || 'general';
+            if (groups[section]) {
+                groups[section].push(opt);
             } else {
                 groups.general.push(opt);
             }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@aws/ml-container-creator",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Generator for SageMaker AI BYOC paradigm for predictive inference use-cases.",
   "type": "module",
   "main": "src/app.js",
@@ -70,10 +70,11 @@
   ],
   "scripts": {
     "test": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --exit --reporter test/reporters/progress-reporter.cjs",
-    "test:ci": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --exit --reporter spec",
+    "test:ci": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --exit --parallel --reporter spec",
     "test:verbose": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --exit --reporter spec",
     "test:minimal": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --reporter test/reporters/minimal-reporter.cjs",
     "test:watch": "VALIDATE_ENV_VARS=false mocha test/**/*.test.js --watch --reporter test/reporters/progress-reporter.cjs",
+    "test:fast": "VALIDATE_ENV_VARS=false PROPERTY_NUM_RUNS=10 mocha test/**/*.test.js --exit --parallel --timeout 60000 --reporter dot",
     "test:coverage": "VALIDATE_ENV_VARS=false nyc mocha test/**/*.test.js",
     "test:unit": "VALIDATE_ENV_VARS=false mocha test/unit/**/*.test.js test/helpers/**/*.test.js --reporter spec",
     "test:integration": "VALIDATE_ENV_VARS=false mocha test/input-parsing-and-generation/**/*.test.js --reporter spec",
@@ -96,6 +97,8 @@
     "docs:deploy": "mkdocs gh-deploy",
     "docs:sync": "node scripts/sync-command-generator.js",
     "docs:check": "node scripts/sync-command-generator.js --check",
+    "codegen": "node scripts/codegen-cli.js && node scripts/codegen-validator.js && node scripts/codegen-widget.js",
+    "codegen:check": "node scripts/codegen-cli.js && node scripts/codegen-validator.js && node scripts/codegen-widget.js && node scripts/codegen-parity.js",
     "_sbom": "npm sbom --sbom-format spdx > sbom.json && for dir in servers/*/; do [ -f \"$dir/package.json\" ] && (cd \"$dir\" && npm sbom --sbom-format spdx > sbom.json); done",
     "_licenses:review": "license-checker --production --exclude MIT,Apache-2.0,BSD-2-Clause,BSD-3-Clause,ISC,0BSD && for dir in servers/*/; do [ -f \"$dir/package.json\" ] && echo \"\\nChecking $dir\" && (cd \"$dir\" && npx license-checker --production --exclude MIT,Apache-2.0,BSD-2-Clause,BSD-3-Clause,ISC,0BSD); done",
     "_licenses:csv": "license-checker --csv --out ./licenses.csv && for dir in servers/*/; do [ -f \"$dir/package.json\" ] && (cd \"$dir\" && npx license-checker --csv --out licenses.csv); done && cat servers/*/licenses.csv >> licenses.csv",

package/src/app.js

@@ -296,6 +296,11 @@ export async function writeProject(templateDir, destDir, answers, registryConfig
     // Build ignore patterns
     const ignorePatterns = [];
 
+    // EJS partials — included by templates at render time, not copied to output
+    ignorePatterns.push('**/serve.d/**');
+    ignorePatterns.push('**/deploy.d/**');
+    ignorePatterns.push('**/clean.d/**');
+
     if (answers.deploymentTarget !== 'hyperpod-eks') {
         ignorePatterns.push('**/hyperpod/**');
     }

package/src/lib/config-manager.js

@@ -26,6 +26,7 @@ import DeploymentConfigResolver from './deployment-config-resolver.js';
 import BootstrapConfig from './bootstrap-config.js';
 import { parseKeyValue } from './key-value-parser.js';
 import ParameterSchemaValidator from './parameter-schema-validator.js';
+import { validationRules } from './generated/validation-rules.js';
 
 const __configMgrFilename = fileURLToPath(import.meta.url);
 const __configMgrDir = dirname(__configMgrFilename);
@@ -2261,6 +2262,22 @@ export default class ConfigManager {
      * @private
      */
     _validateParameterValue(parameter, value, context = {}) {
+        // First pass: schema-derived validation rules (type, range, pattern, enum)
+        // Skip deprecated params — they have relaxed validation handled by the switch below
+        const schemaRule = validationRules[parameter];
+        if (schemaRule && value !== null && value !== undefined) {
+            // Don't apply strict enum validation to internally-derived values
+            // The switch statement below handles context-dependent validation
+            const skipSchemaValidation = ['framework', 'modelServer', 'deploymentConfig'].includes(parameter);
+            if (!skipSchemaValidation) {
+                const error = schemaRule(value);
+                if (error) {
+                    throw new ValidationError(error, parameter, value);
+                }
+            }
+        }
+
+        // Second pass: context-dependent validations that require runtime state
         const supportedOptions = this._getSupportedOptions();
         
         switch (parameter) {