@aicgen/aicgen 1.1.0 → 1.2.0

package/.gitmodules

@@ -1,3 +1,4 @@
 [submodule "data"]
 	path = data
-	url = https://github.com/lpsandaruwan/aicgen-docs
+	url = git@github.com:aicgen/aicgen-data.git
+	branch = main

package/AGENTS.md

@@ -4,7 +4,8 @@
 
 **Language:** typescript
 **Type:** cli
-**Architecture:** event-driven
+**Instruction Level:** full
+**Architecture:** layered
 
 ## Development Guidelines
 
@@ -16,11 +17,17 @@ This project follows structured coding guidelines across multiple categories:
   - async
   - interfaces-types
 
+### Architecture
+
+  - boundaries
+  - communication
+  - data
+
 ### Testing
 
   - unit-fundamentals
   - unit-mocking
-  - basics
+  - integration
 
 ### Security
 
@@ -31,6 +38,7 @@ This project follows structured coding guidelines across multiple categories:
 ### Performance
 
   - basics
+  - caching
   - async
 
 ### API Design
@@ -49,15 +57,10 @@ This project follows structured coding guidelines across multiple categories:
   - strategy
   - basics
 
-### Architecture
-
-  - principles
-  - messaging
-  - patterns
-
 ### DevOps
 
   - ci-cd
+  - practices
   - observability
 
 ### Best Practices
@@ -69,8 +72,8 @@ This project follows structured coding guidelines across multiple categories:
 ### Design Patterns
 
   - base-patterns
+  - concurrency
   - data-access
-  - domain-logic
 
 
 ## Commands
@@ -101,12 +104,28 @@ npm run build
 See tool-specific instruction files for detailed code style guidelines:
 - Claude Code: `CLAUDE.md`
 - GitHub Copilot: `.github/copilot-instructions.md`
-- Gemini: `.gemini/instructions.md`
 - Antigravity: `.agent/rules/instructions.md`
+- OpenAI Codex: `.codex/instructions.md`
 
 ## Architecture
 
-This project follows **event-driven** architecture. See architecture guidelines in tool-specific files.
+This project follows **layered** architecture. See architecture guidelines in tool-specific files.
+
+## Agentic Profile
+
+**Profile level:** full
+
+Enabled surfaces:
+- Universal agent instructions: AGENTS.md (passive, stable)
+- Codex skills: plugins/aicgen-sdlc/skills/*/SKILL.md (agentic, stable)
+- Project-local plugin: plugins/aicgen-sdlc/.codex-plugin/plugin.json (agentic, stable)
+- Lifecycle hooks: .codex/hooks.json (side-effecting, stable)
+- MCP templates: .aicgen/mcp/codex.md (side-effecting, stable)
+
+Limits and gated surfaces:
+- No additional gated capabilities for this profile.
+
+Side-effecting features such as hooks, MCP, plugin setup scripts, or executable tool configuration must stay local, deterministic, and explicitly reviewed before use.
 
 ## Testing
 
@@ -119,6 +138,19 @@ Follow testing guidelines in tool-specific instruction files.
 - Run tests before pushing
 - Keep PRs focused and reviewable
 
+## Workflows
+
+Use the project-local `aicgen-sdlc` Codex plugin for structured SDLC delivery:
+
+Flow: `/aicgen-spec` → `/aicgen-research` → `/aicgen-plan` → `/aicgen-build` → `/aicgen-check` → `/aicgen-ship`
+
+- `/aicgen-spec` — Capture the full specification for a feature or task before any code is written. (legacy aicgen `/spec`)
+- `/aicgen-research` — Analyze the active spec with internal codebase scanning and external web research. (legacy aicgen `/research`)
+- `/aicgen-plan` — Produce a phased, checkpoint-driven implementation plan based on the spec and research findings. (legacy aicgen `/plan`)
+- `/aicgen-build` — Execute the next (or a specified) phase of the current implementation plan. (legacy aicgen `/build`)
+- `/aicgen-check` — Verify the current implementation against the active spec — tests, code review, and regression check. (legacy aicgen `/check`)
+- `/aicgen-ship` — Pre-flight wrap-up — verify everything is ready, then draft a PR description. (legacy aicgen `/ship`)
+
 ---
 
 *Generated by aicgen - Universal AI agent instructions following the AGENTS.md standard*

package/README.md

@@ -7,23 +7,25 @@
 </p>
 
 <p align="center">
-  <img src="https://img.shields.io/badge/version-1.0--beta-cyan" alt="Version" />
+  <img src="https://img.shields.io/badge/version-1.2.0-cyan" alt="Version" />
   <img src="https://img.shields.io/badge/license-MIT-purple" alt="License" />
   <img src="https://img.shields.io/badge/bun-%3E%3D1.0.0-cyan" alt="Bun" />
   <img src="https://img.shields.io/github/actions/workflow/status/aicgen/aicgen/test.yml?branch=main&label=tests" alt="Tests" />
-  <img src="https://img.shields.io/badge/coverage-94%25-brightgreen" alt="Coverage" />
+  <img src="https://img.shields.io/badge/tests-144-brightgreen" alt="Tests" />
 </p>
 
 ---
 
-**aicgen** makes your project AI-ready in seconds. Generate tailored instruction files for your AI coding assistant with an interactive CLI wizard.
+**aicgen** makes your project AI-ready in seconds. Generate tailored instruction files, workflows, skills, and guarded agentic surfaces for the coding tools your team actually uses.
 
 ## ✨ Features
 
-- **🎯 Multi-Assistant Support** - Claude Code, GitHub Copilot, Gemini, Antigravity, Codex
-- **📚 82+ Guidelines** - Organized into 12 categories (Language, Architecture, Testing, Security, etc.)
+- **🎯 Multi-Assistant Support** - Claude Code, GitHub Copilot, Antigravity, Codex
+- **📚 99 Guidelines** - Organized into 12 categories (Language, Architecture, Testing, Security, etc.)
 - **🎨 Interactive CLI** - Professional wizard with smart defaults and back navigation
-- **⚡ Hooks & Sub-Agents** - Auto-generates Claude Code hooks and verification agents
+- **⚡ Agentic Profiles** - Three progressive levels: `basic`, `standard`, and `full`
+- **🔁 SDLC Workflows** - `/spec`, `/research`, `/plan`, `/build`, `/check`, and `/ship`
+- **🧩 Codex SDLC Plugin** - Project-local `aicgen-sdlc` plugin with namespaced `/aicgen-*` skills
 - **🏗️ Architecture Aware** - Supports Microservices, Modular Monoliths, Hexagonal, and more
 - **📦 Zero External Dependencies** - All guideline data is embedded in the binary
 
@@ -39,7 +41,7 @@ The CLI will:
 1.  Detect your project language and structure
 2.  Guide you through assistant, architecture, and detail level selection
 3.  Let you customize which guidelines to include
-4.  Generate the appropriate config files (`.claude/`, `.github/`, `.agent/`, etc.)
+4.  Generate the appropriate config files (`.claude/`, `.github/`, `.agent/`, `.codex/`, `AGENTS.md`, and local plugin files)
 
 ---
 
@@ -89,6 +91,7 @@ sudo rpm -i aicgen_x86_64.rpm
 ```bash
 git clone https://github.com/aicgen/aicgen.git
 cd aicgen
+git submodule update --init --recursive data
 bun install
 bun run build:binary
 bun run start init
@@ -98,7 +101,19 @@ bun run start init
 
 ## 📚 Guideline System
 
-aicgen uses a **modular guideline architecture** with **82+ guidelines** organized into **12 categories**.
+aicgen uses a **modular guideline architecture** with **99 guidelines** organized into **12 categories**. Release builds embed content from the `data/` submodule, or from `AICGEN_DATA_DIR` when developing against a sibling `aicgen-data` checkout.
+
+**Release 1.2.0 stats:**
+
+| Metric | Count |
+|--------|-------|
+| Guidelines | 99 |
+| SDLC workflows | 6 |
+| Categories | 12 |
+| Languages | 12 |
+| Architectures | 10 |
+| Datasources | 2 |
+| Profile levels | `basic`, `standard`, `full` |
 
 ```bash
 # View guideline statistics
@@ -106,7 +121,7 @@ aicgen stats
 ```
 
 **Categories:**
-- **Language** - TypeScript, Python, Go, Rust, Java, C#, Ruby, JavaScript
+- **Language** - TypeScript, JavaScript, Python, Go, Rust, Java, C#, Ruby, Dart, Swift, Kotlin, PHP
 - **Architecture** - Layered, Modular Monolith, Microservices, Event-Driven, Hexagonal
 - **DevOps** - CI/CD, Docker, Observability (Log formats, Metrics)
 - **Best Practices** - SOLID, DRY, Clean Code principles
@@ -114,32 +129,62 @@ aicgen stats
 
 ## 📁 Generated Outputs
 
+Profile levels control how much AICGEN generates:
+
+| Level | Output |
+|-------|--------|
+| `basic` | Stable repo instructions and rules only |
+| `standard` | Instructions plus guided workflows, prompt files, and Codex SDLC skills |
+| `full` | Full agentic setup with skills, subagents, guardrail hooks, plugins, and MCP templates |
+
 ### For Claude Code
 ```text
 CLAUDE.md                      # Master instructions (project root)
 .claude/
-├── settings.json              # Hooks & permissions
+├── settings.json              # Guardrail hooks & permissions
 ├── guidelines/                # Modular guidelines
 │   ├── language.md
 │   ├── architecture.md
 │   └── ...
-└── agents/                    # Sub-agents
-    └── guideline-checker.md
+├── agents/                    # Sub-agents
+│   └── guideline-checker.md
+└── skills/                    # Full profile project skills
 ```
 
 ### For GitHub Copilot
 ```text
 .github/
 ├── copilot-instructions.md    # Master instructions
-└── instructions/              # Topic-specific files
+├── instructions/              # Topic-specific files
+├── prompts/                   # Standard+ reusable prompt files
+└── chatmodes/                 # Full profile VS Code chat modes
 ```
 
-### For Gemini / Antigravity
+### For Antigravity
 ```text
-.gemini/                       # or .agent/
-└── instructions.md            # Consolidated system prompt
+.agent/
+├── rules/
+│   └── instructions.md        # Workspace rule index
+└── workflows/                 # Profile-enabled workflows
 ```
 
+### For Codex
+```text
+AGENTS.md                      # First-class Codex instructions
+.codex/
+├── instructions.md            # Codex development guide
+├── hooks.json                 # Full profile session reminder hook
+└── hooks/                     # Full profile hook scripts
+.agents/
+└── plugins/marketplace.json   # Project-local plugin install entry
+plugins/
+└── aicgen-sdlc/               # SDLC lifecycle plugin
+    ├── .codex-plugin/plugin.json
+    └── skills/aicgen-*/SKILL.md
+```
+
+> Gemini CLI generation has been removed from active assistant targets. Use Antigravity for Google-side agentic coding profiles.
+
 ---
 
 ## 🗺️ Roadmap
@@ -147,16 +192,16 @@ CLAUDE.md                      # Master instructions (project root)
 ### ✅ Completed
 
 - [x] Interactive CLI wizard with back navigation
-- [x] Multi-assistant support (Claude Code, Copilot, Gemini, Antigravity, Codex)
-- [x] 82+ guidelines across 12 categories
+- [x] Multi-assistant support (Claude Code, Copilot, Antigravity, Codex)
+- [x] 99 guidelines across 12 categories
 - [x] Architecture-aware configuration (Layered, Modular Monolith, Microservices, etc.)
-- [x] Auto-generated hooks and sub-agents for Claude Code
+- [x] Profile-gated agentic surfaces for workflows, skills, plugins, hooks, and MCP templates
 - [x] Custom guideline management (add/remove)
 - [x] GitHub-based guideline updates
 
 ### 🚧 Future Enhancements
 
-- [ ] Custom validation hooks
+- [ ] Custom organization policy packs
 - [ ] Guideline versioning and diffing
 - [ ] Project-specific guideline templates
 
@@ -164,27 +209,38 @@ CLAUDE.md                      # Master instructions (project root)
 
 ### Running Tests
 
-The project includes a comprehensive test suite with 60+ tests covering all core functionality:
+The project includes a comprehensive test suite with 144 tests covering all core functionality:
 
 ```bash
+# Check embedded instruction data is fresh
+bun run check:embedded-data
+
+# Validate the data submodule
+data/scripts/validate-data
+
+# Typecheck
+bun run typecheck
+
 # Run all tests
-bun test
+bun run test -- --runInBand
 
 # Run tests with coverage report
-bun test --coverage
+bun run test:coverage -- --runInBand
 
 # Run tests in watch mode
-bun test --watch
+bun run test:watch
 ```
 
 ### Test Coverage
 
-Current test coverage: **94%** (93.44% function coverage, 94.52% line coverage)
+Current release validation: embedded data freshness, data schema validation, typecheck, Jest tests, and Bun build.
 
 **Test Suite Includes:**
 - ✅ GuidelineLoader tests (filtering, level selection, architecture handling)
 - ✅ Tarball extraction tests (CONFIG-based prefix validation)
-- ✅ AssistantFileWriter tests (all 5 assistants - Claude Code, Copilot, Gemini, Antigravity, Codex)
+- ✅ AssistantFileWriter tests (Claude Code, Copilot, Antigravity, Codex)
+- ✅ Agentic capability matrix and profile gating tests
+- ✅ Codex project-local plugin generation tests
 - ✅ File generation and path handling (cross-platform compatibility)
 - ✅ Content validation and metadata inclusion
 
@@ -212,4 +268,4 @@ Contributions are welcome! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for det
 
 ## 📄 License
 
-MIT © 2025
+MIT © 2026

package/claude.md

@@ -1,6 +1,6 @@
 # my-project - Development Guidelines
 
-**Role:** You are an expert software engineer specialized in typescript and event-driven architecture.
+**Role:** You are an expert software engineer specialized in typescript and layered architecture.
 **User's Goal:** Build high-quality, maintainable software following strict project guidelines.
 
 ## Guidelines
@@ -8,13 +8,13 @@
 This project follows structured coding guidelines organized by category:
 
 - **Language**: @.claude/guidelines/language.md
+- **Architecture**: @.claude/guidelines/architecture.md
 - **Testing**: @.claude/guidelines/testing.md
 - **Security**: @.claude/guidelines/security.md
 - **Performance**: @.claude/guidelines/performance.md
 - **API Design**: @.claude/guidelines/api-design.md
 - **Code Style**: @.claude/guidelines/code-style.md
 - **Error Handling**: @.claude/guidelines/error-handling.md
-- **Architecture**: @.claude/guidelines/architecture.md
 - **DevOps**: @.claude/guidelines/devops.md
 - **Best Practices**: @.claude/guidelines/best-practices.md
 - **Design Patterns**: @.claude/guidelines/design-patterns.md
@@ -26,11 +26,40 @@ This project follows structured coding guidelines organized by category:
 - Code style: See Code Style guidelines above
 - Architecture: See Architecture guidelines above
 
+## Workflows
+
+Use these slash commands for structured SDLC delivery:
+
+Flow: `/spec` → `/research` → `/plan` → `/build` → `/check` → `/ship`
+
+- `/spec` — Capture the full specification for a feature or task before any code is written.
+- `/research` — Analyze the active spec with internal codebase scanning and external web research.
+- `/plan` — Produce a phased, checkpoint-driven implementation plan based on the spec and research findings.
+- `/build` — Execute the next (or a specified) phase of the current implementation plan.
+- `/check` — Verify the current implementation against the active spec — tests, code review, and regression check.
+- `/ship` — Pre-flight wrap-up — verify everything is ready, then draft a PR description.
+
+## Agentic Profile
+
+**Profile level:** full
+
+Enabled surfaces:
+- Project memory: CLAUDE.md (passive, stable)
+- Slash commands: .claude/commands/*.md (guided, stable)
+- Project subagents: .claude/agents/*.md (agentic, stable)
+- Project skills: .claude/skills/*/SKILL.md (agentic, stable)
+- Lifecycle hooks: .claude/settings.json (side-effecting, stable)
+
+Limits and gated surfaces:
+- No additional gated capabilities for this profile.
+
+Side-effecting features such as hooks, MCP, plugin setup scripts, or executable tool configuration must stay local, deterministic, and explicitly reviewed before use.
+
 ## Important Notes
 
 - **Follow the guidelines** referenced above.
 - **Verification**: Use sub-agents in `.claude/agents/` to verify compliance.
-- **Constraints**: Hooks in `.claude/settings.json` enforce critical rules.
+- **Constraints**: Hooks in `.claude/settings.json` provide deterministic guardrails; review them before extending.
 
 ---
 *Generated by aicgen*

package/data/README.md

@@ -2,10 +2,22 @@
 
 This repository contains coding guidelines and best practices that power aicgen configurations.
 
+## Release 1.2.0 Stats
+
+| Metric | Count |
+|--------|-------|
+| Guidelines | 99 |
+| SDLC workflows | 6 |
+| Categories | 12 |
+| Languages | 12 |
+| Architectures | 10 |
+| Datasources | 2 |
+| Profile levels | `basic`, `standard`, `full` |
+
 ## Directory Structure
 
 ```
-data/
+aicgen-data/
 ├── guideline-mappings.yml    # Maps guideline IDs to files and filters
 ├── api/                      # API design patterns
 ├── architecture/             # Architecture patterns (clean, DDD, etc.)
@@ -18,10 +30,26 @@ data/
 ├── practices/                # Best practices
 ├── security/                 # Security guidelines
 ├── style/                    # Code style guidelines
+├── agentic/                  # Assistant capability matrix and profile limits
 ├── templates/                # Reusable templates
+├── workflows/                # SDLC lifecycle command definitions
 └── testing/                  # Testing strategies
 ```
 
+## Agentic Profiles
+
+The `agentic/` directory describes how guideline content maps into assistant-specific surfaces for Claude Code, GitHub Copilot, Antigravity, and Codex.
+
+Gemini CLI generation is removed from active targets. Use Antigravity for Google-side agentic coding profiles.
+
+Profile levels gate capability risk:
+
+- `basic`: stable instructions and rules.
+- `standard`: reusable workflows, prompt files, and stable Codex SDLC skills.
+- `full`: focused agents, skills, safe lifecycle guardrail hooks, plugin packaging, and advanced documentation templates such as MCP.
+
+Test-suite execution is handled by the `/check` lifecycle command, not by generated hooks. Hook templates must not run full test suites automatically.
+
 ## Contributing Guidelines
 
 ### Step 1: Create Your Guideline File
@@ -30,7 +58,7 @@ Create a markdown file in the appropriate category folder:
 
 ```bash
 # Example: Adding a new TypeScript guideline
-data/language/typescript/my-guideline.md
+aicgen-data/language/typescript/my-guideline.md
 ```
 
 **Guideline Format:**
@@ -79,7 +107,6 @@ my-guideline-id:
     - typescript
   levels:
     - standard
-    - expert
     - full
   tags:
     - typescript
@@ -93,24 +120,24 @@ my-guideline-id:
 | `path` | Yes | Relative path to the guideline file |
 | `category` | Yes | Display category (Language, Architecture, Testing, etc.) |
 | `languages` | No | Which languages this applies to. Omit for all languages |
-| `levels` | No | Instruction levels: `basic`, `standard`, `expert`, `full`. Omit for all |
+| `levels` | No | Instruction levels: `basic`, `standard`, `full`. Omit for all |
 | `architectures` | No | Architecture types. Omit for all |
 | `tags` | No | Search/organization tags |
 
 ### Available Values
 
 **Languages:**
-- `typescript`, `python`, `go`, `rust`, `java`, `csharp`, `ruby`, `php`, `swift`, `kotlin`
+- `typescript`, `javascript`, `python`, `go`, `rust`, `java`, `csharp`, `ruby`, `dart`, `swift`, `kotlin`, `php`
 
 **Levels:**
 - `basic` - Essential guidelines only
 - `standard` - Common best practices
-- `expert` - Advanced patterns
-- `full` - Comprehensive coverage
+- `full` - Comprehensive coverage with advanced agentic surfaces
 
 **Architectures:**
 - `layered`, `clean-architecture`, `hexagonal`, `ddd`, `microservices`
-- `modular-monolith`, `event-driven`, `serverless`, `other`
+- `modular-monolith`, `event-driven`, `serverless`, `monorepo`
+- `bounded-contexts`, `component-based`
 
 **Categories:**
 - `Language`, `Architecture`, `Testing`, `Security`, `Performance`
@@ -128,7 +155,6 @@ typescript-decorators:
   languages:
     - typescript
   levels:
-    - expert
     - full
   tags:
     - typescript
@@ -144,7 +170,6 @@ solid-principles:
   category: Architecture
   levels:
     - standard
-    - expert
     - full
   tags:
     - solid
@@ -162,7 +187,6 @@ ddd-aggregates:
     - ddd
     - clean-architecture
   levels:
-    - expert
     - full
   tags:
     - ddd
@@ -174,17 +198,31 @@ ddd-aggregates:
 
 After adding a guideline:
 
-1. **Rebuild aicgen:**
+1. **Validate the data repo:**
    ```bash
+   scripts/validate-data
+   ```
+
+2. **Rebuild aicgen:**
+   ```bash
+   AICGEN_DATA_DIR=/path/to/aicgen-data bun run embed
    bun run build
    ```
 
-2. **Check it's loaded:**
+3. **Check it's loaded:**
    ```bash
    bun run start stats
    ```
 
-3. **Test generation:**
+   Expected 1.2.0 level stats:
+
+   ```text
+   basic: 23
+   standard: 83
+   full: 99
+   ```
+
+4. **Test generation:**
    ```bash
    bun run start init --force
    ```

package/data/agentic/README.md

@@ -0,0 +1,22 @@
+# Agentic Capability Profiles
+
+This directory documents how AICGEN maps reusable guideline content into modern AI coding tool surfaces.
+
+Gemini CLI is no longer an active generated target. Use Antigravity for Google-side agentic coding profiles.
+
+## Profile Levels
+
+- `basic`: main repository instructions and stable rule files only.
+- `standard`: basic output plus reusable workflows, commands, and prompt files.
+- `full`: standard output plus focused agents, skills, guardrail hooks, plugin packaging, and advanced integration templates such as MCP.
+
+Validation stays in the lifecycle commands: `/check` is responsible for running or asking for the relevant test suite. Generated hooks should never invoke full test suites automatically.
+
+## Risk Levels
+
+- `passive`: instructions or rules only.
+- `guided`: reusable prompts, commands, or workflows that users invoke.
+- `agentic`: subagents, skills, or plugin-packaged behavior selected by the assistant.
+- `side-effecting`: hooks, MCP, setup scripts, or executable tool configuration.
+
+Side-effecting capabilities must stay local, deterministic, and explicitly reviewed before use.