create-sdd-project 0.2.1 → 0.2.4

package/README.md

@@ -1,18 +1,67 @@
 # SDD DevFlow
 
-**Spec-Driven Development workflow template for AI-assisted coding.**
+[![npm version](https://img.shields.io/npm/v/create-sdd-project)](https://www.npmjs.com/package/create-sdd-project)
+[![license](https://img.shields.io/npm/l/create-sdd-project)](LICENSE)
+[![node](https://img.shields.io/node/v/create-sdd-project)](package.json)
 
-A development methodology designed for Claude Code and Gemini that combines specialized AI agents, workflow orchestration with human checkpoints, and institutional memory. Built for creating robust, maintainable, and scalable TypeScript projects.
+**Spec-Driven Development workflow for AI-assisted coding.**
+
+A complete development methodology for Claude Code and Gemini that combines specialized AI agents, workflow orchestration with human checkpoints, and institutional memory. Works with new and existing TypeScript/JavaScript projects.
+
+## Quick Start
+
+### New Project
+
+```bash
+npx create-sdd-project my-app
+```
+
+The interactive wizard asks about your stack, AI tools, and autonomy level. For defaults (fullstack Express+Next.js):
+
+```bash
+npx create-sdd-project my-app --yes
+```
+
+### Existing Project
+
+```bash
+cd your-existing-project
+npx create-sdd-project --init
+```
+
+Scans your project, detects your stack and architecture, and installs SDD files adapted to your project. Never modifies existing code or overwrites existing files.
+
+### After Setup
+
+Open in your AI coding tool and run:
+
+```
+init sprint 0
+start task B0.1
+```
+
+The workflow skill guides you through each step with checkpoints based on your autonomy level.
+
+---
 
 ## What is SDD?
 
 SDD DevFlow combines three proven practices:
 
 1. **Spec-Driven Development** — Write specifications before code. Specs are the contract between planning and implementation.
-2. **Test-Driven Development** — Red-Green-Refactor cycle for every feature. Tests define expected behavior before implementation.
-3. **Human-in-the-Loop** — Strategic checkpoints (spec, ticket, plan, commit, merge) with configurable autonomy levels that reduce human intervention as trust increases.
+2. **Test-Driven Development** — Red-Green-Refactor cycle for every feature.
+3. **Human-in-the-Loop** — Strategic checkpoints with configurable autonomy levels that reduce human intervention as trust increases.
+
+### Why use SDD DevFlow?
+
+- **AI agents work better with structure.** Without guardrails, AI coding assistants produce inconsistent results. SDD provides the methodology, standards, and workflow that make AI output predictable and high-quality.
+- **Institutional memory across sessions.** Sprint trackers, bug logs, and decision records survive context compaction and session boundaries.
+- **Scales from solo to team.** Start at L1 (full control) while learning, scale to L4 (full auto) for repetitive tasks.
+- **Works with your stack.** Not opinionated about frameworks — detects and adapts to Express, Fastify, NestJS, Next.js, Nuxt, Vue, Angular, and many more.
+
+---
 
-## What's Inside
+## What's Included
 
 ### 9 Specialized Agents
 
@@ -26,35 +75,33 @@ SDD DevFlow combines three proven practices:
 | `qa-engineer` | Edge cases, spec verification | 5 |
 | `database-architect` | Schema design, optimization | Any |
 
+### 3 Skills (Slash Commands)
+
+| Skill | Trigger | What it does |
+|-------|---------|-------------|
+| `development-workflow` | `start task B0.1`, `next task`, `init sprint N` | Orchestrates the complete 7-step workflow |
+| `bug-workflow` | `report bug`, `fix bug`, `hotfix needed` | Bug triage, investigation, and resolution |
+| `project-memory` | `set up project memory`, `log a bug fix` | Maintains institutional knowledge |
+
 ### Workflow (Steps 0–6)
 
 ```
-0. SPEC      → spec-creator drafts specs        → Spec Approval (Std/Cplx)
-1. SETUP     → Branch, ticket, sprint tracker    → Ticket Approval (Std/Cplx)
-2. PLAN      → Planner creates implementation plan → Plan Approval (Std/Cplx)
+0. SPEC      → spec-creator drafts specs        → Spec Approval
+1. SETUP     → Branch, ticket, sprint tracker    → Ticket Approval
+2. PLAN      → Planner creates implementation plan → Plan Approval
 3. IMPLEMENT → Developer agent, TDD
 4. FINALIZE  → Tests/lint/build, validator       → Commit Approval
 5. REVIEW    → PR, code review, QA              → Merge Approval
 6. COMPLETE  → Clean up, update tracker
 ```
 
-**Step flow by complexity:**
-- **Simple**: 1 → 3 → 4 → 5 → 6
-- **Standard**: 0 → 1 → 2 → 3 → 4 → 5 (+QA) → 6
-- **Complex**: 0 → 1 (+ADR) → 2 → 3 → 4 → 5 (+QA) → 6
-
-### 3 Complexity Tiers
-
-| Tier | Spec | Ticket | Plan | QA |
-|------|:----:|:------:|:----:|:--:|
-| Simple | Skip | Skip | Skip | Skip |
-| Standard | Yes | Yes | Yes | Yes |
-| Complex | Yes | Yes + ADR | Yes | Yes |
+**By complexity:**
+- **Simple** (bug fixes, small tweaks): 1 → 3 → 4 → 5 → 6
+- **Standard** (features): 0 → 1 → 2 → 3 → 4 → 5 (+QA) → 6
+- **Complex** (architectural changes): 0 → 1 (+ADR) → 2 → 3 → 4 → 5 (+QA) → 6
 
 ### 4 Autonomy Levels
 
-Control how many human approval checkpoints are active:
-
 | Level | Name | Human Checkpoints | Best For |
 |-------|------|-------------------|----------|
 | L1 | Full Control | All 5 | First sprint, learning SDD |
@@ -64,13 +111,6 @@ Control how many human approval checkpoints are active:
 
 Quality gates (tests, lint, build, validators) **always run** regardless of level.
 
-### Branching Strategy
-
-Configurable per-project in `key_facts.md`:
-
-- **GitHub Flow** (default): `main` + `feature/*` + `hotfix/*`. Best for MVPs.
-- **GitFlow** (scaled): `main` + `develop` + `feature/*` + `release/*` + `hotfix/*`. For larger projects.
-
 ### Project Memory
 
 Tracks institutional knowledge across sessions in `docs/project_notes/`:
@@ -82,70 +122,39 @@ Tracks institutional knowledge across sessions in `docs/project_notes/`:
 
 ### Automated Hooks (Claude Code)
 
-- **Quick Scan** — After developer agents finish, a fast grep-based scan (~2s, no API calls) checks for debug code, secrets, and TODOs. Critical issues block; warnings pass through.
-- **Compaction Recovery** — After context compaction, injects a reminder to read the sprint tracker for context recovery.
-- **Notifications** — Personal notification hooks (macOS/Linux) in `.claude/settings.local.json`.
+- **Quick Scan** — After developer agents finish, a fast grep-based scan (~2s, no API calls) checks for debug code, secrets, and TODOs
+- **Compaction Recovery** — After context compaction, injects a reminder to read the sprint tracker for context recovery
 
 ### Multi-Tool Support
 
-- **Claude Code**: Full support with agents, skills, hooks, and settings (`.claude/`)
-- **Gemini**: Adapted agent format (`.gemini/`), same methodology via `ai-specs/specs/`
-- **Other tools**: `AGENTS.md` provides universal instructions readable by 21+ AI coding tools
+| Tool | Support Level |
+|------|--------------|
+| **Claude Code** | Full — agents, skills, hooks, settings (`.claude/`) |
+| **Gemini** | Full — agents, skills, commands, settings (`.gemini/`) |
+| **Other AI tools** | `AGENTS.md` provides universal instructions (Cursor, Copilot, Windsurf, etc.) |
 
-## Quick Start
+---
 
-### Option A: New Project (recommended)
+## `--init` Stack Detection
 
-```bash
-npx create-sdd-project my-app
-```
-
-The interactive wizard guides you through:
+When running `--init` on an existing project, the scanner automatically detects:
 
-1. **Project basics** — name, description, business context
-2. **Tech stack** — backend (Express+Prisma+PG, Express+MongoDB, or custom), frontend (Next.js+Tailwind or custom), or both
-3. **Configuration** — AI tools (Claude/Gemini/both), autonomy level (L1-L4), branching strategy
-
-Non-interactive mode with defaults:
-```bash
-npx create-sdd-project my-app --yes
-```
-
-### Option B: Existing Project
-
-Add SDD DevFlow to a project that already has code:
-
-```bash
-cd your-existing-project
-npx create-sdd-project --init
-```
-
-The `--init` flag:
-- **Scans** your project: detects stack (Express, Next.js, Prisma, etc.), architecture (MVC, DDD, feature-based), tests, and existing docs
-- **Adapts** standards files to match your real architecture (not generic DDD defaults)
-- **Imports** existing OpenAPI/Swagger specs and references Prisma schemas
-- **Audits** test coverage and suggests retrofit testing tasks if coverage is low
-- **Never** modifies your existing code or overwrites existing files
-
-### After setup
-
-```
-# Open in your AI coding tool and run:
-init sprint 0
-start task B0.1
-```
+| Category | Detected |
+|----------|----------|
+| **Backend frameworks** | Express, Fastify, Koa, NestJS, Hapi, AdonisJS |
+| **ORMs** | Prisma, Mongoose, TypeORM, Sequelize, Drizzle, Knex, MikroORM, Objection.js |
+| **Databases** | PostgreSQL, MySQL, SQLite, MongoDB, SQL Server, CockroachDB (from Prisma schema, `.env`, or dependencies) |
+| **Frontend frameworks** | Next.js, Nuxt, Remix, Astro, SolidJS, React, Vue, Angular, Svelte |
+| **Styling** | Tailwind CSS, styled-components, Emotion, Sass |
+| **Component libraries** | Radix UI, Headless UI, Material UI, Chakra UI, Ant Design |
+| **State management** | Zustand, Redux, Jotai, TanStack Query, Recoil, Pinia, MobX |
+| **Testing** | Jest, Vitest, Mocha (unit) + Playwright, Cypress (e2e) |
+| **Architecture** | MVC, DDD, feature-based, handler-based, flat |
+| **Project type** | Monorepo (workspaces, Lerna, Turbo, pnpm) or single-package |
 
-The workflow skill will guide you through each step with checkpoints based on your autonomy level.
+Standards files are adapted to match your actual architecture — not generic defaults.
 
-### Manual Setup (alternative)
-
-If you prefer manual configuration, copy the template directly:
-
-```bash
-cp -r template/ /path/to/your-project/
-```
-
-Then look for `<!-- CONFIG: ... -->` comments in the files to customize.
+---
 
 ## Template Structure
 
@@ -155,7 +164,6 @@ project/
 ├── CLAUDE.md                            # Claude Code config (autonomy, recovery)
 ├── GEMINI.md                            # Gemini config (autonomy)
 ├── .env.example                         # Environment variables template
-├── .gitignore                           # Git ignore with secrets protection
 │
 ├── .claude/
 │   ├── agents/                          # 9 specialized agents
@@ -165,45 +173,37 @@ project/
 │   │   ├── bug-workflow/                # Bug triage and resolution
 │   │   └── project-memory/              # Memory system setup
 │   ├── hooks/quick-scan.sh              # Post-developer quality scan
-│   ├── settings.json                    # Shared hooks (git-tracked)
-│   └── settings.local.json              # Personal hooks (gitignored)
+│   └── settings.json                    # Shared hooks (git-tracked)
 │
 ├── .gemini/
 │   ├── agents/                          # 9 agents (Gemini format)
-│   ├── skills/
-│   │   ├── development-workflow/        # Main task workflow (Steps 0-6)
-│   │   │   └── references/              # Templates, guides, examples
-│   │   ├── bug-workflow/                # Bug triage and resolution
-│   │   └── project-memory/              # Memory system setup
+│   ├── skills/                          # Same 3 skills
 │   ├── commands/                        # Slash command shortcuts
-│   ├── settings.json                    # Gemini configuration
-│   └── styles/default.md               # Response style
+│   └── settings.json                    # Gemini configuration
 │
 ├── ai-specs/specs/
 │   ├── base-standards.mdc              # Constitution + methodology
-│   ├── backend-standards.mdc           # Backend patterns (DDD, Express, Prisma)
-│   ├── frontend-standards.mdc          # Frontend patterns (Next.js, Tailwind, Radix)
-│   └── documentation-standards.mdc     # Documentation update rules
+│   ├── backend-standards.mdc           # Backend patterns
+│   ├── frontend-standards.mdc          # Frontend patterns
+│   └── documentation-standards.mdc     # Documentation rules
 │
 └── docs/
     ├── project_notes/                   # Project memory
-    │   ├── sprint-0-tracker.md          # Sprint tracker template
-    │   ├── key_facts.md                 # Project configuration
-    │   ├── bugs.md                      # Bug log
-    │   └── decisions.md                 # ADRs
-    ├── specs/
-    │   ├── api-spec.yaml                # OpenAPI spec (backend)
-    │   └── ui-components.md             # Component spec (frontend)
-    └── tickets/                         # Task tickets (generated by workflow)
+    │   ├── sprint-0-tracker.md
+    │   ├── key_facts.md
+    │   ├── bugs.md
+    │   └── decisions.md
+    ├── specs/                           # API and UI specs
+    └── tickets/                         # Task tickets (workflow-generated)
 ```
 
 ## Default Tech Stack
 
-The template defaults to (configurable via `<!-- CONFIG -->` comments):
+Configurable via the wizard or `<!-- CONFIG -->` comments in template files:
 
 - **Backend**: Node.js + Express + Prisma + PostgreSQL
 - **Frontend**: Next.js (App Router) + Tailwind CSS + Radix UI + Zustand
-- **Shared Types**: Zod schemas in `shared/` workspace → `z.infer<>` for TypeScript types
+- **Shared Types**: Zod schemas with `z.infer<>` for TypeScript types
 - **Testing**: Jest (unit) + Playwright (e2e)
 - **Methodology**: TDD + DDD + Spec-Driven Development
 
@@ -224,6 +224,16 @@ These 6 principles apply to ALL tasks, ALL agents, ALL complexity levels:
 - Node.js 18+
 - `jq` (for quick-scan hook): `brew install jq` (macOS) or `apt install jq` (Linux)
 
+## Manual Setup (Alternative)
+
+If you prefer manual configuration over the CLI wizard:
+
+```bash
+cp -r template/ /path/to/your-project/
+```
+
+Then look for `<!-- CONFIG: ... -->` comments in the files to customize.
+
 ## Roadmap
 
 - **Agent Teams**: Parallel execution of independent tasks (waiting for Claude Code Agent Teams to stabilize)

package/lib/config.js

@@ -79,7 +79,7 @@ const AI_TOOLS = [
 
 const AUTONOMY_LEVELS = [
   { level: 1, name: 'Full Control', desc: 'Human approves every checkpoint (first sprint, learning SDD)' },
-  { level: 2, name: 'Trusted', desc: 'Human reviews plans + merges only (default, normal development)', default: true },
+  { level: 2, name: 'Trusted', desc: 'Human reviews plans + merges only (normal development)', default: true },
   { level: 3, name: 'Autopilot', desc: 'Human only approves merges (well-defined, repetitive tasks)' },
   { level: 4, name: 'Full Auto', desc: 'No human checkpoints, CI/CD gates only (bulk simple tasks)' },
 ];

package/lib/init-generator.js

@@ -305,16 +305,16 @@ function adaptBackendStandards(template, scan) {
     '<!-- TODO: Review and adjust the sections below to match your project\'s conventions. -->\n<!-- This file was generated from project analysis by create-sdd-project --init. -->'
   );
 
-  // Update globs in frontmatter
+  // Update globs in frontmatter — only include tsx/jsx if frontend detected
   const srcRoot = findSrcRootName(scan);
-  const globPattern = srcRoot ? `${srcRoot}/**/*.{ts,js,tsx,jsx}` : '**/*.{ts,js,tsx,jsx}';
+  const exts = scan.frontend.detected ? 'ts,js,tsx,jsx' : 'ts,js';
+  const globPattern = srcRoot ? `${srcRoot}/**/*.{${exts}}` : `**/*.{${exts}}`;
   content = content.replace(
     /globs: \[.*?\]/,
     `globs: ["${globPattern}"]`
   );
 
   // Update Technology Stack
-  const framework = scan.backend.framework || 'Unknown';
   const orm = scan.backend.orm;
   const db = scan.backend.db;
   const lang = scan.language === 'typescript' ? 'TypeScript' : 'JavaScript';
@@ -325,8 +325,10 @@ function adaptBackendStandards(template, scan) {
 
   let stackLines = [
     `- **Runtime**: Node.js with ${lang}`,
-    `- **Framework**: ${framework}`,
   ];
+  if (scan.backend.framework) {
+    stackLines.push(`- **Framework**: ${scan.backend.framework}`);
+  }
   if (orm) {
     stackLines.push(`- **ORM**: ${orm}${db ? ` (${db})` : ''}`);
   } else if (db) {
@@ -352,21 +354,38 @@ function adaptBackendStandards(template, scan) {
   const patternLabel = patternLabels[scan.srcStructure.pattern] || 'Custom';
 
   const archStructure = buildArchitectureTree(scan);
+  // Robust: match any "## Architecture — <anything>" heading up to "## Naming"
   content = content.replace(
-    /## Architecture — DDD Layered\n\n```\n[\s\S]*?```\n\n### Layer Rules\n\n[\s\S]*?(?=\n## Naming)/,
+    /## Architecture — [^\n]+\n\n```\n[\s\S]*?```\n\n(?:### Layer Rules\n\n[\s\S]*?)?(?=\n## Naming)/,
     `## Architecture — ${patternLabel}\n\n\`\`\`\n${archStructure}\`\`\`\n\n<!-- TODO: Add layer rules that match your project's architecture. -->\n\n`
   );
 
   // Update Database Patterns section if not Prisma
+  // Robust: match from "## Database Patterns" to next "## " heading
   if (!scan.backend.orm) {
     content = content.replace(
-      /## Database Patterns\n\n### Prisma Best Practices[\s\S]*?### Repository Pattern[\s\S]*?```\n/,
-      `## Database Patterns\n\n<!-- TODO: Add database access patterns for your project. -->\n\n`
+      /## Database Patterns\n\n[\s\S]*?(?=\n## )/,
+      `## Database Patterns\n\n<!-- TODO: Add database access patterns for your project. -->\n`
     );
   } else if (scan.backend.orm !== 'Prisma') {
     content = content.replace(
-      /## Database Patterns\n\n### Prisma Best Practices[\s\S]*?### Repository Pattern[\s\S]*?```\n/,
-      `## Database Patterns\n\n<!-- TODO: Add ${scan.backend.orm} best practices and patterns for your project. -->\n\n`
+      /## Database Patterns\n\n[\s\S]*?(?=\n## )/,
+      `## Database Patterns\n\n<!-- TODO: Add ${scan.backend.orm} best practices and patterns for your project. -->\n`
+    );
+  }
+
+  // Adapt Validation section based on detected validation library
+  if (scan.backend.validation && scan.backend.validation !== 'Zod') {
+    // Has a validator, but not Zod — replace with detected library name
+    content = content.replace(
+      /## Validation\n\n[\s\S]*?(?=\n## Database Patterns)/,
+      `## Validation\n\n- Validate all inputs at the application layer using ${scan.backend.validation}\n- Validate before executing business logic\n- Return descriptive validation errors\n\n<!-- TODO: Add ${scan.backend.validation} validation patterns for your project. -->\n\n`
+    );
+  } else if (!scan.backend.validation) {
+    // No validation library detected — generic guidance
+    content = content.replace(
+      /## Validation\n\n[\s\S]*?(?=\n## Database Patterns)/,
+      `## Validation\n\n- Validate all inputs at the application layer\n- Validate before executing business logic\n- Return descriptive validation errors\n\n<!-- TODO: Add validation patterns for your project (e.g., Zod, Joi, class-validator). -->\n\n`
     );
   }
 
@@ -444,7 +463,7 @@ function buildArchitectureTree(scan) {
 }
 
 function capitalizeFramework(name) {
-  const map = { jest: 'Jest', vitest: 'Vitest', mocha: 'Mocha' };
+  const map = { jest: 'Jest', vitest: 'Vitest', mocha: 'Mocha', playwright: 'Playwright', cypress: 'Cypress' };
   return map[name] || name;
 }
 
@@ -467,15 +486,17 @@ function adaptAgentsMd(template, config, scan) {
   const tree = rootDirs.map((d) => `├── ${d.replace(/\/$/, '/')}   `).join('\n');
   const treeBlock = `\`\`\`\nproject/\n${tree}\n└── docs/        ← Documentation\n\`\`\``;
 
+  // Robust: flexible whitespace between CONFIG comment and code block
   content = content.replace(
-    /<!-- CONFIG: Adjust directories.*?-->\n\n```\nproject\/\n[\s\S]*?```/,
+    /<!-- CONFIG: Adjust directories[^>]*-->\n+```\nproject\/\n[\s\S]*?```/,
     treeBlock
   );
 
   // If not monorepo, simplify the install instructions
+  // Robust: match any number of table rows (not hardcoded count)
   if (!scan.isMonorepo) {
     content = content.replace(
-      /\*\*Critical\*\*: NEVER install dependencies in the root directory\.\n\n\| Action.*\n\|.*\n\|.*\n\|.*\n\|.*\n/,
+      /\*\*Critical\*\*: NEVER install dependencies in the root directory\.\n\n(\|.*\n)+/,
       ''
     );
   }
@@ -494,6 +515,12 @@ function adaptAgentsMd(template, config, scan) {
       'Frontend patterns (Next.js, Tailwind, Radix)',
       `Frontend patterns (${parts.join(', ')})`
     );
+  } else {
+    // Remove frontend-standards reference for backend-only projects
+    content = content.replace(
+      /- \[Frontend Standards\].*\n/,
+      ''
+    );
   }
 
   return content;
@@ -528,9 +555,11 @@ function configureKeyFacts(template, config, scan) {
 
   // Technology Stack from scan
   if (scan.backend.detected) {
-    const fw = scan.backend.framework || 'Unknown';
     const runtime = scan.language === 'typescript' ? 'Node.js (TypeScript)' : 'Node.js';
-    content = content.replace('[Framework, runtime, version]', `${fw}, ${runtime}`);
+    const backendLabel = scan.backend.framework
+      ? `${scan.backend.framework}, ${runtime}`
+      : runtime;
+    content = content.replace('[Framework, runtime, version]', backendLabel);
 
     if (scan.backend.db) {
       content = content.replace('[Type, host, port]', scan.backend.db);

package/lib/init-wizard.js

@@ -33,6 +33,7 @@ function formatScanSummary(scanResult) {
   const patternLabels = {
     mvc: 'MVC',
     ddd: 'DDD (Domain-Driven Design)',
+    layered: 'Layered (controllers + handlers + managers)',
     'feature-based': 'Feature-based',
     'handler-based': 'Handler-based',
     flat: 'Flat structure',

package/lib/prompts.js

@@ -47,29 +47,29 @@ function askMultiline(rl, question) {
     console.log('  (Enter text below. Empty line to finish, or press Enter to skip)');
     const lines = [];
     let firstLine = true;
+    let done = false;
 
-    const onLine = (line) => {
+    const handler = (line) => {
+      if (done) return;
       if (firstLine && line.trim() === '') {
-        rl.removeListener('line', onLine);
+        done = true;
+        rl.removeListener('line', handler);
         resolve('');
         return;
       }
       firstLine = false;
       if (line.trim() === '') {
-        rl.removeListener('line', onLine);
+        done = true;
+        rl.removeListener('line', handler);
         resolve(lines.join('\n'));
       } else {
         lines.push(line);
+        process.stdout.write('  > ');
       }
     };
 
     process.stdout.write('  > ');
-    rl.on('line', (line) => {
-      if (lines.length > 0 || line.trim() !== '') {
-        process.stdout.write('  > ');
-      }
-      onLine(line);
-    });
+    rl.on('line', handler);
   });
 }
 

package/lib/scanner.js

@@ -67,11 +67,12 @@ function detectBackend(dir, pkg) {
 
   // Framework detection
   const frameworks = [
-    { key: 'express', dep: 'express', label: 'Express' },
-    { key: 'fastify', dep: 'fastify', label: 'Fastify' },
-    { key: 'koa', dep: 'koa', label: 'Koa' },
-    { key: 'nestjs', dep: '@nestjs/core', label: 'NestJS' },
-    { key: 'hapi', dep: '@hapi/hapi', label: 'Hapi' },
+    { dep: 'express', label: 'Express' },
+    { dep: 'fastify', label: 'Fastify' },
+    { dep: 'koa', label: 'Koa' },
+    { dep: '@nestjs/core', label: 'NestJS' },
+    { dep: '@hapi/hapi', label: 'Hapi' },
+    { dep: '@adonisjs/core', label: 'AdonisJS' },
   ];
 
   for (const fw of frameworks) {
@@ -84,11 +85,14 @@ function detectBackend(dir, pkg) {
 
   // ORM detection
   const orms = [
-    { key: 'prisma', dep: '@prisma/client', label: 'Prisma' },
-    { key: 'mongoose', dep: 'mongoose', label: 'Mongoose' },
-    { key: 'typeorm', dep: 'typeorm', label: 'TypeORM' },
-    { key: 'sequelize', dep: 'sequelize', label: 'Sequelize' },
-    { key: 'drizzle', dep: 'drizzle-orm', label: 'Drizzle' },
+    { dep: '@prisma/client', label: 'Prisma' },
+    { dep: 'mongoose', label: 'Mongoose' },
+    { dep: 'typeorm', label: 'TypeORM' },
+    { dep: 'sequelize', label: 'Sequelize' },
+    { dep: 'drizzle-orm', label: 'Drizzle' },
+    { dep: 'knex', label: 'Knex' },
+    { dep: '@mikro-orm/core', label: 'MikroORM' },
+    { dep: 'objection', label: 'Objection.js' },
   ];
 
   for (const orm of orms) {
@@ -117,6 +121,22 @@ function detectBackend(dir, pkg) {
   // Port detection
   result.port = detectPort(dir, pkg);
 
+  // Validation library detection
+  const validators = [
+    { dep: 'zod', label: 'Zod' },
+    { dep: 'joi', label: 'Joi' },
+    { dep: 'class-validator', label: 'class-validator' },
+    { dep: 'yup', label: 'Yup' },
+    { dep: 'ajv', label: 'Ajv' },
+  ];
+  result.validation = null;
+  for (const v of validators) {
+    if (deps[v.dep]) {
+      result.validation = v.label;
+      break;
+    }
+  }
+
   // If no framework but has a server-like structure, still mark as detected
   if (!result.detected && (result.orm || result.db)) {
     result.detected = true;
@@ -129,9 +149,13 @@ function detectFrontend(dir, pkg) {
   const deps = getAllDeps(pkg);
   const result = { detected: false, framework: null, styling: null, components: null, state: null };
 
-  // Framework detection
+  // Framework detection (order matters — more specific first)
   const frameworks = [
     { dep: 'next', label: 'Next.js' },
+    { dep: 'nuxt', label: 'Nuxt' },
+    { dep: '@remix-run/react', label: 'Remix' },
+    { dep: 'astro', label: 'Astro' },
+    { dep: 'solid-js', label: 'SolidJS' },
     { dep: 'react', label: 'React' },
     { dep: 'vue', label: 'Vue' },
     { dep: '@angular/core', label: 'Angular' },
@@ -155,6 +179,8 @@ function detectFrontend(dir, pkg) {
   // Component libraries
   if (deps['@radix-ui/react-dialog'] || deps['@radix-ui/react-select'] || hasRadixDep(deps)) {
     result.components = 'Radix UI';
+  } else if (deps['@headlessui/react']) {
+    result.components = 'Headless UI';
   } else if (deps['@mui/material']) {
     result.components = 'Material UI';
   } else if (deps['@chakra-ui/react']) {
@@ -166,6 +192,8 @@ function detectFrontend(dir, pkg) {
   // State management
   if (deps['zustand']) result.state = 'Zustand';
   else if (deps['@reduxjs/toolkit'] || deps['redux']) result.state = 'Redux';
+  else if (deps['jotai']) result.state = 'Jotai';
+  else if (deps['@tanstack/react-query']) result.state = 'TanStack Query';
   else if (deps['recoil']) result.state = 'Recoil';
   else if (deps['pinia']) result.state = 'Pinia';
   else if (deps['mobx']) result.state = 'MobX';
@@ -280,13 +308,14 @@ function detectTests(dir, pkg) {
   const deps = getAllDeps(pkg);
   const result = {
     framework: 'none',
+    e2eFramework: null,
     hasConfig: false,
     testFiles: 0,
     testDirs: [],
     estimatedCoverage: 'none',
   };
 
-  // Framework detection
+  // Unit test framework detection
   if (deps['jest'] || deps['@jest/core'] || deps['ts-jest'] || deps['@types/jest']) {
     result.framework = 'jest';
   } else if (deps['vitest']) {
@@ -295,11 +324,21 @@ function detectTests(dir, pkg) {
     result.framework = 'mocha';
   }
 
+  // E2E test framework detection (supplement, doesn't override unit framework)
+  result.e2eFramework = null;
+  if (deps['@playwright/test'] || deps['playwright']) {
+    result.e2eFramework = 'playwright';
+  } else if (deps['cypress']) {
+    result.e2eFramework = 'cypress';
+  }
+
   // Config files
   const configFiles = [
     'jest.config.js', 'jest.config.ts', 'jest.config.mjs', 'jest.config.cjs',
     'vitest.config.js', 'vitest.config.ts', 'vitest.config.mjs',
     '.mocharc.yml', '.mocharc.json', '.mocharc.js',
+    'playwright.config.ts', 'playwright.config.js',
+    'cypress.config.ts', 'cypress.config.js',
   ];
   result.hasConfig = configFiles.some((f) => fs.existsSync(path.join(dir, f)));
 
@@ -417,7 +456,10 @@ function detectDatabaseFromPrisma(dir) {
     if (fs.existsSync(schemaPath)) {
       try {
         const content = fs.readFileSync(schemaPath, 'utf8');
-        const match = content.match(/provider\s*=\s*"(\w+)"/);
+        // Match provider only within a datasource block (skip generator blocks)
+        const dsBlock = content.match(/datasource\s+\w+\s*\{[^}]*\}/);
+        const providerSource = dsBlock ? dsBlock[0] : content;
+        const match = providerSource.match(/provider\s*=\s*"(\w+)"/);
         if (match) {
           const provider = match[1];
           const dbMap = {
@@ -444,6 +486,7 @@ function detectDatabaseFromEnv(dir) {
     if (fs.existsSync(envPath)) {
       try {
         const content = fs.readFileSync(envPath, 'utf8');
+        // Check DATABASE_URL
         const match = content.match(/DATABASE_URL\s*=\s*(\S+)/);
         if (match) {
           const url = match[1].replace(/["']/g, '');
@@ -452,6 +495,10 @@ function detectDatabaseFromEnv(dir) {
           if (url.startsWith('mysql://')) return 'MySQL';
           if (url.includes('sqlite')) return 'SQLite';
         }
+        // Check MONGODB_URI / MONGO_URI
+        if (/^MONGO(?:DB)?_URI\s*=/m.test(content)) return 'MongoDB';
+        // Check REDIS_URL
+        if (/^REDIS_URL\s*=/m.test(content)) return 'Redis';
       } catch { /* ignore */ }
     }
   }
@@ -466,7 +513,7 @@ function detectPort(dir, pkg) {
     if (fs.existsSync(envPath)) {
       try {
         const content = fs.readFileSync(envPath, 'utf8');
-        const match = content.match(/^PORT\s*=\s*(\d+)/m);
+        const match = content.match(/^PORT\s*=\s*["']?(\d+)/m);
         if (match) return parseInt(match[1], 10);
       } catch { /* ignore */ }
     }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "create-sdd-project",
-  "version": "0.2.1",
+  "version": "0.2.4",
   "description": "Create a new SDD DevFlow project with AI-assisted development workflow",
   "bin": {
     "create-sdd-project": "bin/cli.js"