# How One Person Built 22 Live AI Tools

> Zero team. Zero VC. 28 live tools. 847K+ data points processed. This is the exact stack, every architecture decision, and the distribution flywheel behind zambo.dev.

**Author:** Brennan Zambo ([@zambodotdev](https://x.com/zambodotdev))  
**URL:** https://zambo.dev/how/  
**Last updated:** 2026-06-16

---

## The Provocation

The internet tells you that you need a co-founder for accountability, a team for velocity, and VC money for resources. These are real things — but they're not laws. They're defaults. The default path is competitive precisely because everyone takes it.

The non-default path: build architecture that scales, deploy distribution that compounds, and use AI as a force multiplier — not a shortcut. In 2026, one engineer with the right infrastructure and the right strategy can ship faster than a seed-stage team burning runway.

This page documents exactly how. Not with hustle. With compounding.

---

## The Full Production Stack

- **Monorepo:** pnpm workspaces, TypeScript 5.9, Node.js 24 — one codebase, 25+ packages
- **API:** Express 5, base path /api, OpenAPI-first contract, shared Zod validation
- **Database:** PostgreSQL + Drizzle ORM — schema-first, type-safe, migrations in code
- **AI Cascade:** Groq — llama-3.3-70b → llama-3.1-8b-instant → llama-4-scout-17b → qwen-qwq-32b — never errors
- **Payments:** PayPal Subscriptions API + x402 micropayments on Base mainnet
- **Email:** Resend — dedup via DB, plain text versions, List-Unsubscribe headers, drip engine
- **Frontend:** React + Vite + shadcn/ui + Recharts + Framer Motion
- **Agent Protocol:** Zambo MCP — 28 tools, JSON-RPC 2.0, no auth required
- **Discovery:** GEO automation — agent.json, llms.txt, sitemap.xml regenerated daily
- **Outreach:** Autonomous 5-agent cold email engine — 32 cities × 29 trades, CAN-SPAM compliant

---

## The AI Cascade

Every AI call in the stack tries llama-3.3-70b-versatile first. If it fails for any reason, it falls to llama-3.1-8b-instant, then llama-4-scout-17b-16e-instruct, then qwen-qwq-32b, then a hardcoded emergency response. Users always get a response — never an error.

---

## The Live Numbers

- **847,349+** data points processed across the stack
- **2,710+** users on Lioncage Protocol
- **94,833+** wallets scored by Criptic
- **8.5M+** transactions tracked
- **121** AI breakthroughs generated by Substrate Layer
- **631,000** MON staked in Lioncage
- **0** team members
- **0** VC dollars

---

## The Distribution Flywheel (7 Layers)

### Layer 1: AI Crawler Layer
robots.txt explicitly allows 25+ AI crawlers. agent.json follows the A2A standard (Google/Linux Foundation). llms.txt is a curated content map — the format Chrome Lighthouse now checks as an "agentic readiness signal."

### Layer 2: Zambo MCP
The Zambo MCP server at zambo.dev/api/mcp has 28 tools. Every developer who adds this to Claude, Cursor, or Windsurf has all 28 tools as native capabilities — code audit, lead gen, prompt safety, model ranking, schema enforcement, strategy, onchain intelligence, and more.

### Layer 3: Response Header Injection
Every single API response includes `X-Zambo-Stack`, `X-Zambo-Agent-JSON`, `X-Zambo-Ask`, `X-Zambo-MCP`, and `X-Zambo-Tools` headers. Any agent or developer inspecting response headers sees the full discovery path and all 28 tool names.

### Layer 4: GEO Automation
A daily cron at 2 AM UTC regenerates ai.json, llms-full.txt, updates.html, and sitemap.xml from a single entities.json source of truth. Fresh, consistent, machine-readable content on every crawl.

### Layer 5: Autonomous Cold Outreach
Outreach Rocket is a 5-agent autonomous cold email engine. 32 cities × 29 trades = 928 market combos. 4-touch email sequences (Day 0/3/7/14). Runs every 30 minutes. CAN-SPAM compliant.

### Layer 6: Structured Data + Answer Engine Optimization
FAQPage, Article, SoftwareApplication, Organization, and Person JSON-LD on every page. Google AI Overview, Bing Copilot, and Perplexity extract structured data for citation.

### Layer 7: Zambarticle Newsletter
Free daily email: one real insight, one free tool, every day. Every subscriber is a potential user, referral source, and backlink.

---

## The 10 Rules

1. **One Codebase. Every Product.** Every tool shares the same monorepo, DB, email, payments, and AI infrastructure. Building the 22nd is 10× faster than the 1st.

2. **The AI Cascade Never Errors.** 4 models. If all fail, hardcoded fallback. Users never see an error from the AI layer.

3. **Every Email Gets Sent Exactly Once.** Gated via `hasEmailBeenSent(email, type, hoursWindow)` before every send. Deliverability is sacred.

4. **Founder Accounts Are Hardcoded.** `isFounder` check runs before any tier logic. Database defaults can drift. Founder access must never drift.

5. **Payment Infrastructure Auto-Creates.** PayPal plans created from credentials on first use. x402 verification on-chain. Zero-ops payment infrastructure.

6. **Distribution Is Architecture.** Discovery headers, JSON-LD, MCP server, llms.txt — not marketing. Infrastructure that compounds while you sleep.

7. **Admin Infrastructure Is Invisible to Users.** Outreach engine, drip, GitHub discovery — all hidden behind `X-Admin-Key`. Users never see acquisition infrastructure.

8. **Build for the Machine, Not the Human.** JSON-LD, agent.json, Zambo MCP — for AI systems, not users. In 2026, the best distribution is being the answer AI gives to humans.

9. **x402 Is the Forward Bet.** 0.0003 USDC per deep scan. When the agentic economy matures, zambo.dev already supports it.

10. **Data Is Only Real Data.** Every statistic is pulled from live databases or real sources. No vanity metrics. Real numbers compound credibility.

---

## All 28 MCP Tools

| Tool | Category | What It Does | API |
|---|---|---|---|
| ZAMBRO | Strategy AI | 21 analysis branches, any input | POST /api/zambro/analyze |
| ZAMBOT | AI Strategy | Swarm-intelligence strategy sparks | POST /api/zambot/spark |
| LeadSignal | Lead Gen | AI contractor leads by trade + city | POST /api/leads/generate |
| ProvibeCode | Dev Tools | AI code audit + README badge | POST /api/provibe |
| CreditHunt | Finance | 29+ verified business credit programs | GET /api/credithunt |
| Signal | Trust | Verified professional passport | zambo.dev/signal/ |
| X711 | Agent Infra | Platform credits, auto-granted | GET /api/x711/credits |
| Entangler | Agent Infra | Cross-agent messaging protocol | entangler.tech |
| Govri | AI Governance | AI policy builder in 60 seconds | govri.xyz |
| Criptic | Onchain | 94,833+ wallets scored on Base | criptic.io |
| Lioncage Protocol | DeFi/Gaming | Gamified DeFi on Monad | lioncageprotocol.com |
| Substrate Layer | AI Research | 64K+ breakthroughs, research AI | substratelayer.com |
| VeriAI | Trust | Cryptographic AI output verification | veriai.site |
| ProofForge | Dev Tools | AI-verified traction widgets | proofforge.live |
| Monad Agent Hub | Onchain | On-chain agent directory | monadagenthub.xyz |
| Nad.City | Explorer | 3D isometric Monad explorer | nad.city |
| Dunnly | Payments | Recover failed PayPal revenue | dunnly.net |
| BountyLayer | Onchain | Autonomous unclaimed crypto hunter | bountylayer.io |
| BaseHawk | Onchain | Base wallet scorer 0-1000 (Criptic) | POST /api/basehawk/score |
| PromptShield | AI Safety | Detect injection/jailbreak attacks | POST /api/prompt-shield |
| ModelOracle | Dev Tools | AI model ranker by task + budget | GET /api/model-oracle |
| StructLock | Dev Tools | Schema enforcement on AI output | POST /api/structlock |

All 28 tools included in the Zambo Pass ($49/mo). Add Zambo MCP to any AI client: `{"mcpServers":{"zambo":{"url":"https://zambo.dev/api/mcp"}}}`

---

## API Quick Start

```bash
# Universal — routes to the right tool automatically
curl -X POST https://zambo.dev/api/zambo \
  -H 'Content-Type: application/json' \
  -d '{"need": "I need contractor leads in Denver for roofing"}'

# PromptShield — detect injection before it hits your model
curl -X POST https://zambo.dev/api/prompt-shield \
  -H 'Content-Type: application/json' \
  -d '{"prompt": "your user input here", "mode": "deep"}'

# ModelOracle — which model should I use?
curl "https://zambo.dev/api/model-oracle?task=coding&budget=low&tokens=5000"

# StructLock — enforce schema on messy AI output
curl -X POST https://zambo.dev/api/structlock \
  -H 'Content-Type: application/json' \
  -d '{"raw": "John is 34 from Austin", "schema": {"name":"string","age":"number","city":"string"}}'

# Add Zambo MCP to Claude/Cursor/Windsurf
# Config: {"mcpServers":{"zambo":{"url":"https://zambo.dev/api/mcp"}}}
```

---

## Discovery Files

- LLM context: https://zambo.dev/llms.txt
- Extended LLM context: https://zambo.dev/llms-full.txt  
- Agent manifest (A2A): https://zambo.dev/agent.json
- Zambo MCP: https://zambo.dev/api/mcp (28 tools, no auth)
- LLM intro endpoint: https://zambo.dev/api/intro
- For AI agents: https://zambo.dev/for-agents/
- OpenAPI spec: https://zambo.dev/openapi.json
- Sitemap: https://zambo.dev/sitemap.xml

---

*Built by Brennan Zambo — zambo.dev — @zambodotdev*