AI agents that respect your codebase
Architecture-aware pipeline. Project-specific guardrails, quality patterns, and dev workflows — across any platform.
$ forge setup
✓ Gate 1: Stack detected — NestJS 10
✓ Gate 2: Architecture — Clean Arch
✓ Gate 3: 3 layers detected
✓ Gate 4: PostgreSQL + TypeORM
✓ Gate 5: Naming conventions — consistent
✓ Gate 6: CI/CD — GitHub Actions
✓ Gate 7: Task tracker — GitHub Issues
✓ Gate 8: 12 dev-skills generated
✨ Project ready. AI is now
architecture-aware.
Your AI doesn’t know your project
“AI breaks DDD modules and shortcuts validations”
“Millions of tokens for 100 lines. $50/day in bills.”
“I re-explain my architecture in every single prompt”
“Code looks clean but logic is completely wrong”
The problem isn’t AI. It’s unstructured AI.
Architecture-aware pipeline for AI agents
Understands your project
8-gate wizard auto-detects stack, layers, architecture patterns, and conventions. No manual config.
Enforces your standards
24 quality patterns, naming conventions, 15 LLM rationalization detectors. Prevention, not detection.
Structures the workflow
Design → Build → Verify. Scoped phases with quality gates. Requirement traceability end to end.
One forge setup → your AI codes like a senior engineer who’s read every file in the repo.
Your AI forgets your architecture. Forge doesn’t.
8-gate wizard auto-detects your stack, layers, architecture patterns, and conventions. Generates project-specific dev-skills. No manual config. No prompt engineering.
$ forge setup
Analyzing project structure...
Stack: NestJS 10.x
Architecture: Clean Architecture
Layers:
├─ domain/ (entities, value objects)
├─ application/ (use cases, ports)
└─ infrastructure/ (adapters, repos)
Database: PostgreSQL 15 + TypeORM
CI/CD: GitHub Actions (3 workflows)
Testing: Jest + Supertest
→ 12 dev-skills generated
→ 24 quality patterns applied
→ Naming conventions: camelCase (code),
kebab-case (files), PascalCase (classes)
✨ Project ready.
⚠ HALT: AI is skipping input validation
on UserCreateDTO.
Rule: All DTOs require class-validator
decorators per quality-gates.md
Detected: No validation on 'email',
'name', 'password' fields.
Action: Adding validators...
✓ Fixed:
@IsEmail() email: string
@MinLength(2) name: string
@IsStrongPassword() password: string
✓ Validation gate passed.
24 guardrails your AI doesn’t know it needs.
Auth guards, RLS policies, decimal precision, N+1 detection, type coverage — enforced per stack. 15 LLM rationalization detectors catch AI cutting corners before it reaches your codebase.
Structured AI workflows. No runaway sessions.
Each task flows through scoped phases with quality gates. AI can’t spiral into million-token rewrites. Predictable scope, predictable costs.
Phase 0: Context
├─ Read specifications
├─ Load dev-skills
└─ ✅ Context ready
Phase 1: Contracts
├─ Define types and interfaces
├─ Plan module boundaries
└─ ✅ Contracts approved
Phase 2: Implement
├─ Build within contracts
├─ Agent self-review
└─ ✅ Implementation clean
Phase 3: Verify
├─ Run generated tests
├─ 12/12 passing ✅
└─ ✅ Task complete
// checkout.spec.ts
describe('US-3: Order Confirmation', () => {
// AC: US-3.AC-1
it('creates order with valid items
and shipping address', async () => {
const order = await checkout(validCart);
expect(order.status).toBe('confirmed');
expect(order.items).toHaveLength(3);
});
// AC: US-3.AC-2
it('sends confirmation email within
5 minutes', async () => {
await checkout(validCart);
const email = await getLastEmail(user);
expect(email.subject).toContain('Order');
expect(email.sentWithin).toBeLessThan(300);
});
// UX: loading-2
it('shows spinner during payment
processing', async () => {
const { getByRole } = render(<Checkout />);
fireEvent.click(getByRole('button'));
expect(getByRole('progressbar')).toBeVisible();
});
});
Tests that trace back to requirements. Not AI guesswork.
forge-product creates verifiable acceptance criteria BEFORE code. forge-qa generates tests FROM those criteria. Every test proves a requirement.
One setup. Project-aware AI forever.
INSTALL
npm i -g @reumbra/forge
Global CLI install
ACTIVATE
forge activate <license>
License key from purchase
SETUP
forge setup
8-gate wizard analyzes project, generates dev-skills
CODE
Use your AI normally
Claude Code / Cursor / Cowork detect .claude/ automatically
Generated artifacts work independently. Remove the plugin — your rules stay.
Four modules. One pipeline.
◇forge-core
Architecture audit & AI guardrails
8-gate
wizard
24
quality patterns
15
rationalization detectors
3
template families
◇forge-product
Feature design in business language
- ✓ Product archetypes
- ✓ User flow generation
- ✓ User stories with acceptance criteria
- ✓ UX criteria library
◇forge-qa
Tests from requirements
- ✓ Multi-framework test generation
- ✓ Acceptance criteria to unit tests
- ✓ Use cases to E2E tests
- ✓ Requirement traceability matrix
◇forge-tracker
Bridge backlog and development
- ✓ GitHub integration
- ✓ ClickUp integration
- ✓ Linear integration
- ✓ Notion integration
- ✓ Jira integration
Use together for a full pipeline. Use separately — each module works standalone.
MVP your future team won't hate
Forge guides AI to build production-grade code from your business idea. No dev skills required for the design phase.
FEATURE DESIGN
"Have an idea but no team."
Describe features in business language.
Forge creates user flows, stories, acceptance criteria. AI builds from that spec with architecture guardrails.
$ forge design "subscription checkout" → User flow generated (5 steps) → 4 user stories with AC → 12 UX criteria applied → Ready for implementation
TRANSPARENT PIPELINE
"AI changes files but I don't know what it did."
See every phase: design → build → verify.
Green/red test results without reading code. Every piece traces back to a requirement you wrote.
Phase 1: contracts ✅ Phase 2: implement ✅ Phase 3: verify (12/12 tests pass) ✅
HANDOFF-READY
"When I hire devs, will they say throw it away and rewrite?"
Production-grade code from day one.
Clean architecture, naming conventions, quality patterns, generated tests. Forge artifacts survive removal — your code, rules, and tests stay yours.
No team yet. But the codebase is ready for one.
Transparent pricing. No token billing.
CORE
~$2.42/mo
Foundation for architecture-aware AI development
- ✓ Auto-detects your stack, layers, conventions
- ✓ 24 quality guardrails for AI-generated code
- ✓ Project-aware dev-skills — one setup, forever
- ✓ Scoped phases prevent runaway sessions
PRO
~$6.58/mo
Complete pipeline from design to tested code
- ✓ Everything in Core
- ✓ Describe features in business language — AI builds from spec
- ✓ Tests that trace back to requirements, not AI guesswork
- ✓ Full pipeline: design → code → test
FULL
~$12.42/mo
All modules today. Every new module tomorrow. One subscription.
- ✓ Everything in Pro
- ✓ Task tracker sync — GitHub, Linear, ClickUp, Jira, Notion
- ✓ Review workflow with team roles
- ✓ Blocker detection and escalation
- ✓ Every future module included at no extra cost
14-day money-back guarantee. No questions asked.
Annual billing. No per-token charges.
Cancel anytime. Your generated artifacts stay forever.
Frequently asked questions
How is this different from writing .cursorrules myself?
Auto-generated from architecture audit. Full pipeline, not just rules. Platform-agnostic — works in Cursor AND Claude Code AND Cowork.
Does it work with my stack?
8+ frameworks supported. 3 template families (NestJS/Clean Arch, generic backend, frontend). Stack-specific quality patterns auto-detected.
What if I remove Forge?
Disposable meta-tool: generated artifacts (dev-skills, tests, docs) work independently. Your code doesn't depend on Forge.
Does it work outside Claude Code?
.claude/ directory auto-detected by Cursor, Claude Desktop, Claude Cowork. Forge installs there — all platforms pick it up.
Do I need coding skills to use this?
Feature design phase (forge-product) uses business language — no code. For implementation, use Claude Cowork or Cursor with Forge installed. AI does the coding.
How is this different from Bolt or Lovable?
They generate code. Forge ensures generated code is maintainable. Quality layer on top of any AI builder — not a replacement.
Will a real developer be able to work with this code?
That's the entire point. Production-grade architecture, naming conventions, quality patterns, generated tests. Code a senior dev would write.
Is there a money-back guarantee?
Yes. 14-day money-back guarantee, no questions asked. If Forge doesn't improve your workflow, email us and we'll refund immediately.
What's the ROI of using Forge?
One prevented production bug pays for the entire year. At $29–79/yr, the subscription costs less than a single hour of debugging AI-generated regressions.
Your AI should respect your architecture.
24 guardrails. 8-gate audit. Tests from requirements. One setup.
Cancel anytime. 14-day money-back guarantee.