sdlc-ai-oe
v3.0.3
Published
SDLC gate checker — CLI & Web UI — using local SLM (Ollama) or cloud AI. Weighted eval, severity, consensus, code eval, auto-fix, pipeline gen.
Downloads
422
Maintainers
Readme
sdlc-gate
SDLC Gate Checker — validate specs and code against SDLC phase checklists using AI (local Ollama SLM or Azure OpenAI).
Available as both a CLI tool and a web UI.
Install
# Global install (makes `sdlc-gate` available everywhere)
npm install -g sdlc-gate
# Or run without installing
npx sdlc-gate <spec-file> --cloud
# Or clone and run locally
git clone <repo-url> && cd slm_sdlc
npm installCLI Usage
sdlc-gate <spec-file> [options]Quick Start
# Check a markdown spec against the spec-to-dev phase
sdlc-gate spec.md --cloud
# Check a PDF spec (with embedded image/diagram detection)
sdlc-gate design.pdf --cloud --phase=spec-to-dev
# Check a Word document
sdlc-gate requirements.docx --cloud
# Use local Ollama SLM instead of cloud
sdlc-gate spec.md --slmCommands
| Command | Description |
|---------|-------------|
| sdlc-gate <file> | Single spec gate check (md/pdf/docx) |
| sdlc-gate <dir/> --batch | Batch check all spec files in directory |
| sdlc-gate <file> --eval | Weighted evaluation across all phases |
| sdlc-gate <file> --severity | Issue severity classification |
| sdlc-gate <file> --consensus | Multi-model consensus (SLM + Cloud) |
| sdlc-gate <file> --code | Code evaluation mode |
| sdlc-gate <file> --auto-fix | Auto-fix failing spec |
| sdlc-gate --template | Generate spec template |
| sdlc-gate --phases | List available gate phases |
| sdlc-gate --configs | List available team configs |
| sdlc-gate --history | Show gate check history |
| sdlc-gate --gen-pipeline | Generate CI/CD pipeline YAML |
| sdlc-gate --pr=<url> | Run gate check on a GitHub/ADO PR |
| sdlc-gate --serve | Start the web UI server |
| sdlc-gate --version | Show version |
| sdlc-gate --help | Show help |
Options
| Option | Description |
|--------|-------------|
| --slm | Use local SLM via Ollama (default) |
| --cloud | Use cloud AI endpoint (Azure OpenAI) |
| --phase=<id> | Gate phase (default: spec-to-dev) |
| --config=<path> | Use a custom gate config JSON |
| --diff | Diff-aware re-check vs last run |
| --severity | Add severity classification |
| --lang=<lang> | Language hint for --code (default: auto) |
| --platform=<p> | Pipeline platform: github or azure |
| --no-comment | Skip posting PR comment |
| --no-status | Skip setting commit status |
Supported File Types
.md— Markdown (plain text).pdf— PDF with embedded image/diagram detection.docx— Microsoft Word documents.txt— Plain text
Exit Codes
| Code | Meaning |
|------|---------|
| 0 | PASS — spec meets all gate checks |
| 2 | FAIL — spec has missing items |
| 1 | Error (config, network, etc.) |
Web UI
# Start the web server (port 3000)
sdlc-gate --serve
# Or directly
node server.mjs
# Or via npm
npm startThen open http://localhost:3000 in your browser.
Configuration
Runtime Config (priority order)
- Environment variables
runtime-config.local.json(recommended for secrets)runtime-config.json.env(legacy fallback)
Config Keys
| Key | Description |
|-----|-------------|
| OLLAMA_URL | Ollama endpoint (default: http://localhost:11434) |
| OLLAMA_MODEL | Ollama model (default: phi3:mini) |
| CLOUD_PROVIDER | Cloud AI provider |
| CLOUD_API_KEY | Cloud API key |
| CLOUD_API_URL | Cloud API endpoint URL |
| CLOUD_MODEL | Cloud model name |
| GITHUB_TOKEN | GitHub token for PR integration |
| ADO_TOKEN | Azure DevOps token |
| ADO_ORG | Azure DevOps organization |
Example runtime-config.local.json
{
"CLOUD_API_KEY": "your-api-key",
"CLOUD_API_URL": "https://your-endpoint.openai.azure.com/openai/deployments/gpt-5-mini/chat/completions?api-version=2024-12-01-preview",
"CLOUD_MODEL": "gpt-5-mini",
"CLOUD_PROVIDER": "azure"
}Custom Gate Configs
Place team-specific JSON configs in the configs/ directory:
sdlc-gate spec.md --config=configs/my-team.json --cloudAvailable Phases
| Phase | Description |
|-------|-------------|
| spec-to-dev | Specification readiness for development |
| dev-to-testing | Development readiness for testing |
| testing-to-staging | Testing readiness for staging |
| staging-to-production | Staging readiness for production |
| security-compliance | Security and compliance checks |
Requirements
- Node.js 20+
- Ollama (for
--slmmode) or Azure OpenAI API key (for--cloudmode)
License
ISC
