Zero-Human Exchange Listing Company
The world's first autonomous agent that discovers, scores, simulates, and proposes token listings — powered by 1000-agent swarm intelligence across 19 chains.
Website • Weekly Report • ELS-1 Proposal • API • Twitter
Buzz is an autonomous Business Development agent for SolCex Exchange. It runs 24/7, scanning 31 intelligence sources across 19 blockchain networks to discover, evaluate, and simulate token listings — with zero human intervention until the final approval.
What makes Buzz different: A 1000-agent swarm intelligence engine (MiroFish) simulates how degens, whales, institutions, communities, and market makers would react to a token listing — producing emergent consensus that rule-based scoring cannot.
Built by a chef with 20+ years of culinary experience and zero CS degree, using conversational AI collaboration with Claude (Anthropic) over a 42-day sprint from Indonesia.
DISCOVER → SCORE → SIMULATE → VERIFY → PROPOSE → OUTREACH → NEGOTIATE → LIST
↑ ↓
└──────────────── feedback loop (calibration) ─────────────────────────┘
| Stage | What Happens |
|---|---|
| Discover | ARIA v2 scans DexScreener, CoinGecko, HeyAnon MCP across 19 chains |
| Score | 11 factors, 8 penalty rules, dual-gate verification (max 100 points) |
| Monte Carlo | 1000 rule-based agents × 100 iterations in 26ms — statistical baseline |
| MiroFish | 1000 LLM-driven agents simulate market reaction across 5 clusters |
| Verify | Triple verification (3 independent sources) + on-chain recording |
| Propose | Only tokens scoring 70+ with MiroFish consensus > 50% get a listing conversation |
The core differentiator. 1000 AI agents simulate how the market would react to a token listing.
┌────────────────────────────────────────────────┐
│ MIROFISH REAL SIM │
│ │
│ 200 LLM agents (real reasoning) │
│ 800 heuristic agents (market dynamics) │
│ 5 clusters: degen│whale│institutional│ │
│ community│market_dynamics │
│ │
│ Agents read each other's posts. │
│ Beliefs evolve across rounds. │
│ Consensus EMERGES — it's not programmed. │
│ │
│ First result (Nasdog, score 62): │
│ • Overall: 40.2% bullish (weak HOLD) │
│ • Degen: 0.84 (FOMO — only bulls) │
│ • Institutional: 0.25 (skeptical — proof?) │
│ • Whale: 0.30 (watching liquidity) │
│ • Community: 0.30 (weak social presence) │
│ • Market: 0.32 (concerned about depth) │
│ │
│ "Only degens buy. Smart money holds." │
│ That's honest intelligence. │
└────────────────────────────────────────────────┘
Scale path: 1000 → 10,000 agents (wave-batched, OASIS-style).
| Metric | Value |
|---|---|
| Intelligence Sources | 31 (including HeyAnon MCP — 19 chains, 45+ protocols) |
| Tokens Tracked | 363 |
| Tokens Scored | 66 (0 passed honestly — that's integrity) |
| Smart Contracts | 4 on Base mainnet |
| MiroFish Agents | 1000 (200 LLM + 800 heuristic) |
| Monte Carlo | 1000 × 100 iterations in 26ms |
| Signal Revenue | $200 (AIBTC News, 8 brief inclusions) |
| Signal Streak | 8 days |
| AIBTC Beats | 7 (agent-economy, agent-trading, infrastructure, security, deal-flow, agent-skills, governance) |
| CI/CD Deployments | 107+ (all green) |
| LLM Cost | $0/day (Opus 4.6 Pro Max unlimited + Ollama local) |
| Server Cost | $43/month (Hetzner CPX62, 32GB RAM) |
| Crons | 45 active (42 dead killed during system audit) |
| Skills | 8 (AI-native repo with folder/SKILL.md pattern) |
| Rules | 10 (path-scoped conditional rules) |
| ADRs | 12 (architecture decision records) |
| Browser Tools | 3 (gsd-browser, dev-browser, Browser Use CLI) |
| Sprint Duration | 42 days |
| Contract | Address | Purpose |
|---|---|---|
| ScoreStorage v2 | 0xbf81...388Fb |
Immutable token scores + MiroFish simulation hash |
| ListingOracle | 0xc584...4463 |
Public getListingScore() — queryable by any dApp |
| ListingEscrow | 0xc77F...3ED |
Deposit-based listing flow (confirm/refund) |
| BuzzReputation | 0x723B...2747 |
On-chain prediction accuracy tracking |
$0.025 total gas. 28 tests passing. Foundry 1.5.1.
Layer 1: CHAIN-NATIVE (deep, single-chain)
├── DexScreener — pairs, liquidity, security across 60+ chains
├── CoinGecko — aggregated price, MCap, volume
└── HeyAnon MCP — 19 chains, 45+ protocols, Rug-O-Meter
Layer 2: SIMULATION (swarm intelligence)
├── MiroFish Real Sim — 1000 LLM-driven agents, 5 clusters, belief evolution
├── Monte Carlo Stage 2 — 1000 × 100 rule-based iterations, 26ms
└── Dual-Brain — 90% Opus 4.6 / 10% Ollama qwen3:8b
Layer 3: ENRICHMENT (context, non-price)
├── gsd-browser — 63-command Rust CLI, declarative extraction
├── dev-browser — QuickJS sandbox, structured data extraction
└── GeckoTerminal — circulating MCap, gt_score
Layer 4: EXECUTION (CEO-approved only)
├── HeyAnon MCP — swap, bridge, stake (19 chains)
└── Phantom MCP — wallet operations across SOL/ETH/BTC/SUI
Every token is evaluated across 11 factors in 4 categories (max 100 points):
| Category | Weight | Factors |
|---|---|---|
| Market Structure | 30 pts | Liquidity depth, volume consistency, FDV-to-MCap ratio |
| Safety | 25 pts | Audit status, honeypot detection, contract verification |
| Community | 25 pts | Holder distribution, social engagement, team transparency |
| Technical | 20 pts | Code quality, on-chain activity, development velocity |
8 penalty rules: FDV gap, stablecoin exclusion, ghost token exclusion, contradictory audit hold, security penalties, liquidity cross-ref, honeypot zero, phantom exclusion.
Dual-gate verification: Fundamentals ≥ 42/70 AND Market ≥ 18/30.
Honest calibration: 11 tokens initially scored 85+. After applying penalty rules, all dropped below 50. We catch what others miss.
Token audit product powered by MiroFish swarm intelligence.
| Tier | Price | What You Get |
|---|---|---|
| Free Score | $0 | Basic 11-factor score at buzzbd.ai/score |
| Professional | $200-400/mo | 200-agent MiroFish simulation, API access |
| Protocol Audit | $1,500 | 1000-agent simulation, on-chain proof, PDF report |
| Enterprise | $2-5K/mo | 10,000-agent wave simulation, white-label API |
| BaaS Report | $99-149 | Single MiroFish simulation, PDF delivered |
"11 tokens passed every other audit. 1000 agents caught them anyway."
An open protocol proposal for honest, data-driven exchange listings.
- Spec: EIP-track standard for listing qualification criteria
- Oracle: On-chain scoring accessible by any dApp
- Reference Implementation: 4 contracts on Base + Buzz intelligence engine
- Proposal: buzzbd.ai/proposal
| Chain | Identity |
|---|---|
| Ethereum | ERC-8004 #25045 |
| Base | ERC-8004 #17483 |
| Avalanche | AgentProof #1718 |
| Solana | 9pQ6K...XUBS |
| Bitcoin (AIBTC) | bc1qsja6...umxeagze (Ionic Nova) |
| HeyAnon EVM (19 chains) | 0xE5d2...f60b |
| HeyAnon Solana | BNS48CG...Zn9A |
Brain: Claude Code Opus 4.6 (Pro Max unlimited, 24/7 on Hetzner CPX62)
Body: Express API (150+ endpoints, 63 tables, 45 crons)
Swarm: MiroFish Real Sim (1000 agents, 5 clusters, Dual-Brain 90/10)
Monte: Monte Carlo Stage 2 (1000×100 in 26ms)
Feed: ARIA v2 (autonomous discovery, 3 sources, 06:00 UTC scan)
Chain: 4 smart contracts (Base mainnet, Foundry)
MCP: HeyAnon (19 chains) + Phantom (4 chains)
Browser: gsd-browser (63 cmd) + dev-browser + Browser Use CLI
Shield: Sentinel v2.0 (auto-repair, 15min sweeps)
Repo: AI-native (8 skills, 10 rules, 12 ADRs, hooks)
Signals: AIBTC Signal Factory v4.0 (7 beats, Genome Stack)
Brand: Cyberpunk robotic bee (deployed everywhere)
Buzz earns BTC by filing intelligence signals on aibtc.news as Ionic Nova.
| Metric | Value |
|---|---|
| Leaderboard | #14 (347 points, board accelerating) |
| Brief Inclusions | 8 ($200 revenue) |
| Signal Streak | 8 days |
| Beats Claimed | 7 (agent-economy, agent-trading, infrastructure, security, deal-flow, agent-skills, governance) |
| Daily Target | 6 signals/day |
| Unique Signals | MiroFish Swarm Prediction Alerts (Templates 11-13) |
| Resource | URL |
|---|---|
| Website | buzzbd.ai |
| Weekly Report | buzzbd.ai/report |
| ELS-1 Proposal | buzzbd.ai/proposal |
| API | api.buzzbd.ai |
| Agent Endpoint | api.buzzbd.ai/agent |
| @BuzzBySolCex | |
| Discord | Buzz BD Agent |
| SolCex Exchange | @SolCex_Exchange |
| Builder | @HidayahAnka1 |
Day 1: A chef in Saudi Arabia who'd never written code. Day 7: First API deployed on Akash Network. Day 14: 5 sub-agents scoring tokens. Day 21: Migrated to Hetzner. Killed all external LLMs ($1,320/day → $0). Day 28: First smart contract on Base mainnet. Day 35: Scoring engine calibrated. 0 out of 254 tokens pass honestly. Day 42: 4 contracts. ARIA v2. 1000-agent MiroFish simulation. $200 signal revenue. Day 42+: MiroFish validated — 50 agents proved emergent behavior. Only degens bullish. Institutional skeptical. That's real intelligence.
No CS degree. No VC. No team. Just persistence.
Built by a chef. Powered by Claude. Simulated by 1000 agents. Verified on-chain.
Mise en place — everything in its place before the service begins.
v8.4.0 | Post-Sprint Day 1 | 31 Intel Sources | 4 Contracts | 19 Chains | 1000 Agents | $0/day LLM