⚡ Forge SDLC

Universal Capability-Oriented SDLC Framework & Autonomous Multi-Agent Orchestrator
Run the best agentic engineering skills across BMAD, GitHub Spec Kit, and Internal Engines through a single unified CLI and native IDE slash commands.

🎯 The Problem Forge Solves

Modern agentic coding ecosystems force developers into single-vendor silos:

• Use BMAD and you're locked into BMAD commands and prompts.
• Use Spec Kit and you're locked into Spec Kit pipelines.

In reality, no single framework is best at everything:

• BMAD excels in specialized C4 Technical Architecture, Deep Ambiguity Elicitation, and 5-Lens Code Reviews (bmad-review).
• Spec Kit excels in Spec-Driven Development (spec.md), Project Constitutions (constitution.md), Technical Plans (plan.md), Atomic Tasks (tasks.md), Consistency Audits (analysis.md), and Readiness Convergence (convergence.md).
• Forge Internal excels in STRIDE Threat Modeling, OWASP SAST Security Audits, Test Suite Synthesis, and KeepAChangelog Packaging.

Forge decouples your intent from vendor tooling.

Instead of learning vendor-specific commands (bmad-architecture, speckit-plan), you simply ask for the capability:

npx forge-sdlc architecture

Forge dynamically scores all candidate providers based on project context, recommends the best tool with full explainability ("Why"), and executes it seamlessly through a unified artifact and functionality pipeline.

🚀 Quickstart (Zero Installation Required)

Run instantly anywhere via npx:

# 1. Run the Complete 14-Stage SDLC Pipeline for a feature:
npx forge-sdlc sdlc --functionality auth-module

# 2. Establish Non-Negotiable Invariants & Constitution:
npx forge-sdlc constitution

# 3. Design C4 Technical Architecture (dynamically routes to BMAD):
npx forge-sdlc architecture

# 4. Generate Given-When-Then Specification (dynamically routes to Spec Kit):
npx forge-sdlc specify

# 5. Run 5-Lens Multi-Perspective Code Review (dynamically routes to BMAD):
npx forge-sdlc review

# 6. Check 100% Pipeline Artifact Progress:
npx forge-sdlc status

Or install globally:

npm install -g forge-sdlc
forge sdlc --functionality billing

🤖 15 Active AI Editor Slash Commands & Agents

Forge automatically configures native slash commands and autonomous agent rules for Cursor, Claude Code, Antigravity / Gemini, GitHub Copilot, and Windsurf:

npx forge-sdlc init
# or
npx forge-sdlc agent-rules

🌟 Active Agents in your AI Chat:

🧠 Universal Brainstorm-First Architecture (All Agents):
Every single agent in Forge operates with brainstorming on top. Before generating code or artifacts, each agent ingests .forge/artifacts/brainstorm.md, explores alternative solutions, evaluates edge cases and architectural trade-offs, and then converges on the optimal result.

| Slash Command | Agent Role | Specialized Provider | Mandatory Ingestion Files | Target Generated Artifact | | :--- | :--- | :---: | :--- | :--- | | /brainstorm | 💡 Stage 0: Brainstorm & Lateral Ideation | BMAD | Problem Statement, Vision | brainstorm.md (Foundation for All Agents) | | /sdlc | 🚀 Full SDLC Master Orchestrator | Multi-Provider | Workspace Context | Complete 15-Stage Lifecycle | | /implement | 💻 Senior Staff Implementation | BMAD / Spec Kit | tasks.md, spec.md, architecture.md | Source Code in src/, implementation.md (Brainstorm on Top) | | /heal | 🩺 Auto-Healing & Drift Sync | Internal | spec.md, architecture.md, tasks.md | healing-plan.md, tasks.md | | /swarm | 🐝 Multi-Provider Swarm Consensus | Multi-Provider | Codebase context | swarm-<cap>.md | | /brd | 📊 Business Requirements & ROI | BMAD | Goals & Context, brainstorm.md | brd.md, discovery.md | | /constitution| 🏛️ Constitution & Invariants | Spec Kit | Project Requirements, brainstorm.md | constitution.md | | /specify | 📐 Given-When-Then Specification| Spec Kit | constitution.md, brd.md, brainstorm.md | spec.md | | /clarify | 🔍 Ambiguity Elicitation | BMAD | spec.md, brainstorm.md | clarifications.md | | /architecture| 🏗️ C4 System Design & ADRs | BMAD | spec.md, constitution.md, brainstorm.md | architecture.md, data-model.md | | /plan | 📅 Phased Milestone Roadmap | Spec Kit | spec.md, architecture.md, brainstorm.md | plan.md | | /tasks | ✅ Atomic Developer Checklist | Spec Kit | plan.md, spec.md, constitution.md, brainstorm.md| tasks.md | | /analyze | 🔬 Cross-Artifact Drift Audit | Spec Kit | spec.md, architecture.md, tasks.md, brainstorm.md| analysis.md | | /test | 🧪 Automated QA & Test Suites | Internal | Source Code, spec.md | test-report.md, tests in tests/ | | /review | 🛡️ 5-Lens Code Review | BMAD (bmad-review) | Source Code, architecture.md, brainstorm.md | review.md | | /security | 🔒 STRIDE & OWASP SAST Scan | Internal | Source Code, package.json, brainstorm.md | security-audit.md, threat-model.md | | /converge | 🎯 Release Readiness Burndown | Spec Kit | tasks.md, test-report.md, review.md| convergence.md | | /release | 📦 KeepAChangelog & SemVer Bump | Internal | convergence.md, Git history | CHANGELOG.md, RELEASE_NOTES.md |

📁 Functionality Folders Architecture

Forge organizes all agent outputs into self-contained Functionality / Feature Modules under .forge/functionalities/<feature>/:

.forge/
├── artifacts/                           # Active workspace artifacts (latest synchronized)
│   ├── brainstorm.md                    # 💡 Lateral Ideation & Feasibility (BMAD)
│   ├── discovery.md
│   ├── brd.md
│   ├── constitution.md
│   ├── spec.md
│   ├── clarifications.md
│   ├── architecture.md
│   ├── plan.md
│   ├── tasks.md
│   ├── analysis.md
│   ├── implementation.md
│   ├── test-report.md
│   ├── review.md
│   ├── security-audit.md
│   ├── convergence.md
│   ├── CHANGELOG.md
│   └── RELEASE_NOTES.md
│
├── functionalities/                     # 📂 DEDICATED FUNCTIONALITY MODULES
│   ├── authentication/                  # Full 15-Agent Suite for 'authentication'
│   │   ├── brainstorm.md
│   │   ├── constitution.md
│   │   ├── spec.md
│   │   ├── architecture.md
│   │   ├── tasks.md
│   │   ├── implementation.md
│   │   ├── manifest.json
│   │   └── workflow-state.json
│   │
│   ├── billing-module/                  # Full 15-Agent Suite for 'billing-module'
│   └── core/                            # Default core functionality suite
│
├── iterations/                          # Historical SDLC iteration snapshots
└── runs/                                # Execution snapshots & execution logs

🔄 The 14-Stage SDLC Pipeline & 16 Artifacts (100% Complete)

1. DISCOVERY (/brd) ────────► 2. CONSTITUTION (/constitution) ────────► 3. SPECIFICATION (/specify)
         │                                                                       │
         ▼                                                                       ▼
6. PLANNING (/plan) ◄──────── 5. ARCHITECTURE (/architecture) ◄──────── 4. CLARIFICATION (/clarify)
         │
         ▼
7. TASKS (/tasks)   ────────► 8. ANALYSIS (/analyze)         ────────► 9. IMPLEMENTATION (/implement)
                                                                                 │
                                                                                 ▼
12. SECURITY (/security) ◄─── 11. REVIEW (/review)            ◄──────── 10. TESTING (/test)
         │
         ▼
13. CONVERGENCE (/converge) ──► 14. RELEASE (/release)

🧮 Intelligent Multi-Factor Scoring

Forge dynamically scores all candidate providers using a multi-factor weighted algorithm:

$$\text{Score} = w_1 \cdot \text{Match} + w_2 \cdot \text{Spec} + w_3 \cdot \text{WorkflowFit} + w_4 \cdot \text{ArtifactFit} + w_5 \cdot \text{Context} + w_6 \cdot \text{Quality} + w_7 \cdot \text{UserPref}$$

View provider scores and explainability for any capability:

npx forge-sdlc recommend
npx forge-sdlc matrix

📦 Programmatic TypeScript / Node.js SDK

Forge is also available as a programmatic SDK:

import { CapabilityRouter, WorkflowEngine, ArtifactManager } from 'forge-sdlc';

// 1. Intelligent Capability Routing
const router = new CapabilityRouter();
const recommendation = await router.recommend('architecture');
console.log(`Recommended Provider: ${recommendation.recommendedProvider.providerName}`);

// 2. Execute with Functionality Isolation
const result = await router.execute({
  capabilityId: 'specify',
  functionality: 'payment-gateway',
});

// 3. Execute 14-Stage SDLC Pipeline
const engine = new WorkflowEngine();
const fullSdlc = engine.getAvailableWorkflows().find(w => w.id === 'full-sdlc');
await engine.executeWorkflow(fullSdlc!, undefined, { functionality: 'payment-gateway' });

📜 License

Distributed under the MIT License. See LICENSE for more information.

Built with ⚡ by Rahulrock-1
Universal Capability-Oriented SDLC Framework