NestJS is a TypeScript workable choice for ai system. GreatCTO auto-detects both — adds the ai-system archetype overlay, wires ai-system-specific gates, and runs 83 specialist agents around your existing NestJS workflow.
GreatCTO reads your package.json and detects nestjs + ai-system archetype from signals: imports, file structure, env vars, README hints.
Attaches the ai-system archetype overlay: EU AI Act + GDPR + OWASP LLM gates, training-data lineage, model card requirements. Override if your specifics differ; the defaults are sensible for NestJS-style projects.
qa-engineer runs tsc --strict / eslint / vitest --coverage; security-officer checks for prototype pollution + XSS sinks; performance-engineer reviews bundle size + cold-start times.
Bugs you've hit before in other NestJS projects (connection-pool exhaustion, ORM N+1 queries, retry storms) — the agent's Step 0 includes the prior detection order. MTTR drops 94 % on second occurrence (methodology).
$ cd my-nestjs-app && npx great-cto init ✓ scanning manifests… found package.json ✓ stack: nestjs (TypeScript) ✓ archetype: ai-system ⚠ archetype + stack combo is unusual — review overlay manually ✓ 83 agents ready $ /start "add model inference endpoint" ▸ architect drafting ARCH-ai-system.md… ▸ pm decomposing into beads tasks… ⚐ gate:plan — your approval needed
Approve → 3 senior-devs run in parallel worktrees → 5 reviewers fan out in parallel → gate:ship → deploy. One real run walked stage-by-stage: /proof.
This is the shape of what senior-dev drafts for "model inference endpoint" — auth first, schema validation, and the audit line the ai-system reviewer requires before gate:ship opens.
// src/ai-system/ai-system.controller.ts — reviewed by 5 agents
@Controller('ai-system')
export class AiSystemController {
@Post()
@UseGuards(AuthGuard) // security-officer: auth before handler
async create(@Body() dto: CreateAiSystemDto, @User() user) {
const result = await this.service.handle(dto, user); // qa: class-validator enforced
await this.audit.log(user.id, 'model inference endpoint', result.confidence); // gate:ai-system
return result;
}
}ai-system overlay.Model-serving APIs with drift detection and shadow deploys.
RAG pipelines with citation grounding and eval suites.
Fine-tuning workflows with training-data lineage.
NestJS (TypeScript) is not a typical fit for ai system. The archetype overlay still attaches, but you may want to override defaults more aggressively. Check the ai-system archetype page for the typical stack list and decide if your case is the right tool / right archetype.
No black-box "AI does it all" loop. GreatCTO is a deterministic state machine — 8 stages, 22 nodes, 2 human gates. Every node maps to a real agent on GitHub. Inspect the state machine →
$ npx great-cto init
Free, MIT, runs locally. Built as a Claude Code plugin — install with one command.
Eight stages, two human gates, four memory layers. Why this exact shape, and what I tried that didn't work.
One run, one feature, from prompt to merged PR. Time, cost, and gate-by-gate breakdown — no marketing math.
Regex vs LLM-based archetype detection, the false-positive count, and why I keep rejecting the obvious fix.
The bottleneck in agentic SDLC isn't model quality — it's process governance. Here's the state machine that closes the gap.