SKILL: sdlc-import
Version: 2.0.0
Status: Active
Category: Automation
Phase: 0 (Preparation)
Agent: sdlc-importer
Purpose
Reverse engineer existing codebases (not necessarily legacy - can be modern projects without SDLC artifacts) and automatically generate complete SDLC Agêntico documentation structure.
Key Capabilities:
- Detect 30 programming languages and frameworks (v2.0.0 - expanded from 10)
- Extract 5-15 architecture decisions with confidence scores
- Generate 3-5 architecture diagrams (Mermaid + DOT)
- Perform STRIDE threat modeling
- Identify 10-50 tech debt items (P0-P3 prioritized)
- Create complete
.project/ structure (respects settings.json configuration)
- Auto-create feature branch before analysis
Design Principles
Hybrid Approach (per ADR-022):
- Pattern Extraction - Static analysis using regex/AST parsing
- LLM Synthesis - Claude Opus for ambiguous decisions (confidence < 0.5)
- Plugin Integration - claude-plugins-official for production-grade analysis
Read-Only Analysis:
- Never modifies source code
- Only creates
.project/ documentation (configurable via settings.json)
Configuration Over Code:
- All thresholds in YAML files
- Extensible pattern libraries
- Output directory respects settings.json configuration
Usage
# Basic usage
/sdlc-import /path/to/project
# With options
/sdlc-import /path/to/project --skip-threat-model --no-llm
# Create GitHub issues for P0 debt
/sdlc-import /path/to/project --create-issues
# Custom branch name
/sdlc-import /path/to/project --branch-name feature/import-my-project
Options:
--skip-threat-model - Skip STRIDE analysis--skip-tech-debt - Skip debt detection--no-llm - Disable LLM synthesis (faster, lower cost)--create-issues - Create GitHub issues for P0 debt and HIGH threats--branch-name - Custom branch name (default: auto-generated)
Components
| Script |
Purpose |
project_analyzer.py |
Main orchestrator |
language_detector.py |
Language/framework detection (LSP plugins) |
decision_extractor.py |
ADR inference engine |
architecture_visualizer.py |
Diagram generation (Mermaid/DOT) |
threat_modeler.py |
STRIDE analysis (security-guidance plugin) |
tech_debt_detector.py |
Tech debt scanner (code-review plugin) |
documentation_generator.py |
Output generation |
confidence_scorer.py |
Confidence calculation |
Outputs
Output Directory Configuration:
The skill respects settings.json configuration for output location.
Priority order:
.claude/settings.json → sdlc.output.project_artifacts_dir (default: .project)import_config.yml → general.output_dir (fallback, deprecated)- Default:
.project
Created Files:
.project/ ← Default (configurable)
├── corpus/nodes/decisions/
│ ├── ADR-INFERRED-001.yml
│ ├── ADR-INFERRED-002.yml
│ └── ...
├── security/
│ └── threat-model-inferred.yml
├── architecture/
│ ├── component-diagram.mmd
│ ├── dependency-graph.dot
│ └── data-flow.mmd
└── reports/
├── tech-debt-inferred.md
└── import-report.md
IMPORTANT: Never write to .agentic_sdlc/ - this directory is for framework files only.
Quality Gate
Gate: sdlc-import-gate.yml
Validation Checks:
- Branch created (feature/import-*)
- Overall confidence >= 0.6
- Minimum 3 ADRs extracted
- Zero CRITICAL threats
- Minimum 2 diagrams generated
Supported Technologies (v2.0.0)
30 languages, frameworks, and tools across multiple categories:
Programming Languages (14):
- Python, JavaScript, TypeScript, Java, C#, Go, Ruby, PHP, Rust, Kotlin
- NEW: C++, Gradle (build scripts), Dart (Flutter), Swift (iOS)
Backend Frameworks:
- Django, Flask, FastAPI, Spring, ASP.NET, Express, NestJS, Gin, Rails, Laravel, Symfony, Actix, Ktor
- NEW: CMake, Conan, vcpkg, Boost, Tokio, async-std
Frontend Frameworks:
- React, Angular, Next.js
- NEW: Vue.js, Svelte, Tailwind CSS
Mobile Frameworks:
- NEW: React Native, Flutter, SwiftUI/UIKit, Jetpack Compose, Xamarin
Infrastructure as Code:
- Terraform, Kubernetes
- NEW: Bicep (Azure), Ansible
Configuration Management:
CI/CD Tools:
- GitHub Actions, GitLab CI
- NEW: Jenkins
Testing Frameworks:
- pytest, jest, junit, mocha
- NEW: Selenium (multi-language), Playwright (multi-language)
Build Tools:
Disambiguation Features:
- Chef vs Ruby detection (via metadata.rb marker)
- Ansible vs YAML detection (via ansible.cfg marker)
- Gradle vs Kotlin detection (file-based)
Integration
claude-plugins-official Plugins:
- LSP Plugins - Deep language analysis
- Original: pyright-lsp, typescript-lsp, jdtls-lsp, csharp-lsp, gopls-lsp, rust-analyzer-lsp
- NEW (v2.0.0): clangd-lsp (C++), dart-lsp (Flutter), sourcekit-lsp (Swift)
- security-guidance - Vulnerability scanning, STRIDE templates
- code-review - Code quality, tech debt detection
- claude-md-management - ADR/README generation
- pr-review-toolkit - GitHub issue creation
awesome-copilot Patterns:
reverse-project-analysis.promptarchitecture-blueprint-generator.promptlanguage-stack-identifier.promptthreat-model-generator.promptcode-gap-audit.prompt
References
Version History
- v1.0.0 (2026-01-23) - Initial implementation
1---2name: sdlc-import3description: SKILL: sdlc-import4---5# SKILL: sdlc-import67**Version:** 2.0.08**Status:** Active9**Category:** Automation10**Phase:** 0 (Preparation)11**Agent:** sdlc-importer1213---1415## Purpose1617Reverse engineer existing codebases (not necessarily legacy - can be modern projects without SDLC artifacts) and automatically generate complete SDLC Agêntico documentation structure.1819**Key Capabilities:**20- Detect **30 programming languages and frameworks** (v2.0.0 - expanded from 10)21- Extract 5-15 architecture decisions with confidence scores22- Generate 3-5 architecture diagrams (Mermaid + DOT)23- Perform STRIDE threat modeling24- Identify 10-50 tech debt items (P0-P3 prioritized)25- Create complete `.project/` structure (respects settings.json configuration)26- Auto-create feature branch before analysis2728---2930## Design Principles3132**Hybrid Approach** (per ADR-022):331. **Pattern Extraction** - Static analysis using regex/AST parsing342. **LLM Synthesis** - Claude Opus for ambiguous decisions (confidence < 0.5)353. **Plugin Integration** - claude-plugins-official for production-grade analysis3637**Read-Only Analysis:**38- Never modifies source code39- Only creates `.project/` documentation (configurable via settings.json)4041**Configuration Over Code:**42- All thresholds in YAML files43- Extensible pattern libraries44- Output directory respects settings.json configuration4546---4748## Usage4950```bash51# Basic usage52/sdlc-import /path/to/project5354# With options55/sdlc-import /path/to/project --skip-threat-model --no-llm5657# Create GitHub issues for P0 debt58/sdlc-import /path/to/project --create-issues5960# Custom branch name61/sdlc-import /path/to/project --branch-name feature/import-my-project62```6364**Options:**65- `--skip-threat-model` - Skip STRIDE analysis66- `--skip-tech-debt` - Skip debt detection67- `--no-llm` - Disable LLM synthesis (faster, lower cost)68- `--create-issues` - Create GitHub issues for P0 debt and HIGH threats69- `--branch-name` - Custom branch name (default: auto-generated)7071---7273## Components7475| Script | Purpose |76|--------|---------|77| `project_analyzer.py` | Main orchestrator |78| `language_detector.py` | Language/framework detection (LSP plugins) |79| `decision_extractor.py` | ADR inference engine |80| `architecture_visualizer.py` | Diagram generation (Mermaid/DOT) |81| `threat_modeler.py` | STRIDE analysis (security-guidance plugin) |82| `tech_debt_detector.py` | Tech debt scanner (code-review plugin) |83| `documentation_generator.py` | Output generation |84| `confidence_scorer.py` | Confidence calculation |8586---8788## Outputs8990**Output Directory Configuration:**9192The skill respects `settings.json` configuration for output location.9394**Priority order:**951. `.claude/settings.json` → `sdlc.output.project_artifacts_dir` (default: `.project`)962. `import_config.yml` → `general.output_dir` (fallback, deprecated)973. Default: `.project`9899**Created Files:**100```101.project/ ← Default (configurable)102├── corpus/nodes/decisions/103│ ├── ADR-INFERRED-001.yml104│ ├── ADR-INFERRED-002.yml105│ └── ...106├── security/107│ └── threat-model-inferred.yml108├── architecture/109│ ├── component-diagram.mmd110│ ├── dependency-graph.dot111│ └── data-flow.mmd112└── reports/113 ├── tech-debt-inferred.md114 └── import-report.md115```116117**IMPORTANT:** Never write to `.agentic_sdlc/` - this directory is for framework files only.118119---120121## Quality Gate122123**Gate:** `sdlc-import-gate.yml`124125**Validation Checks:**126- Branch created (feature/import-*)127- Overall confidence >= 0.6128- Minimum 3 ADRs extracted129- Zero CRITICAL threats130- Minimum 2 diagrams generated131132---133134## Supported Technologies (v2.0.0)135136**30 languages, frameworks, and tools across multiple categories:**137138**Programming Languages (14):**139- Python, JavaScript, TypeScript, Java, C#, Go, Ruby, PHP, Rust, Kotlin140- **NEW:** C++, Gradle (build scripts), Dart (Flutter), Swift (iOS)141142**Backend Frameworks:**143- Django, Flask, FastAPI, Spring, ASP.NET, Express, NestJS, Gin, Rails, Laravel, Symfony, Actix, Ktor144- **NEW:** CMake, Conan, vcpkg, Boost, Tokio, async-std145146**Frontend Frameworks:**147- React, Angular, Next.js148- **NEW:** Vue.js, Svelte, Tailwind CSS149150**Mobile Frameworks:**151- **NEW:** React Native, Flutter, SwiftUI/UIKit, Jetpack Compose, Xamarin152153**Infrastructure as Code:**154- Terraform, Kubernetes155- **NEW:** Bicep (Azure), Ansible156157**Configuration Management:**158- **NEW:** Chef, Puppet159160**CI/CD Tools:**161- GitHub Actions, GitLab CI162- **NEW:** Jenkins163164**Testing Frameworks:**165- pytest, jest, junit, mocha166- **NEW:** Selenium (multi-language), Playwright (multi-language)167168**Build Tools:**169- **NEW:** Vite, Webpack170171**Disambiguation Features:**172- Chef vs Ruby detection (via metadata.rb marker)173- Ansible vs YAML detection (via ansible.cfg marker)174- Gradle vs Kotlin detection (file-based)175176---177178## Integration179180**claude-plugins-official Plugins:**181- **LSP Plugins** - Deep language analysis182 - **Original:** pyright-lsp, typescript-lsp, jdtls-lsp, csharp-lsp, gopls-lsp, rust-analyzer-lsp183 - **NEW (v2.0.0):** clangd-lsp (C++), dart-lsp (Flutter), sourcekit-lsp (Swift)184- **security-guidance** - Vulnerability scanning, STRIDE templates185- **code-review** - Code quality, tech debt detection186- **claude-md-management** - ADR/README generation187- **pr-review-toolkit** - GitHub issue creation188189**awesome-copilot Patterns:**190- `reverse-project-analysis.prompt`191- `architecture-blueprint-generator.prompt`192- `language-stack-identifier.prompt`193- `threat-model-generator.prompt`194- `code-gap-audit.prompt`195196---197198## References199200- **ADR:** ADR-022 (Automated Legacy Project Onboarding)201- **Epic:** Issue #52202- **awesome-copilot:** https://github.com/awesome-copilot/awesome-copilot203- **claude-plugins-official:** https://github.com/anthropics/claude-plugins-official204205---206207## Version History208209- **v1.0.0** (2026-01-23) - Initial implementation
arbgjr