RONIN detects emerging narratives across the Solana ecosystem by cross-referencing on-chain activity, developer signals, and market data. It then generates actionable build ideas for each narrative.
Live Demo: ronin-wheat.vercel.app
API Endpoint: GET /api/analyze returns structured JSON of all detected narratives, signals, and build ideas.
RONIN collects signals from multiple independent data sources, cross-references them to identify patterns, and synthesizes emerging narratives with confidence scores.
Data Collection Analysis Output
─────────────── ────────── ────────
Solana RPC (TPS, ──> Pattern ──> Detected Narratives
slots, epoch) Matching (ranked by confidence)
DeFi Llama (TVL, ──> Cross-Source ──> Supporting Signals
protocol growth) Correlation (with data points)
CoinGecko (prices, ──> Confidence ──> Build Ideas (3-5 per
volume, momentum) Scoring narrative, with
feasibility ratings)
GitHub (new repos, ──> Trend
stars, topics) Classification
Each data source produces signals - discrete observations with quantified strength:
| Source | What It Detects | API |
|---|---|---|
| On-Chain | TVL movements, protocol growth/decline, sector trends, network throughput | DeFi Llama, Solana RPC |
| Market | Price momentum, volume concentration, ecosystem token performance | CoinGecko |
| GitHub | Developer activity by topic, new project velocity, trending repositories | GitHub API |
Signals are matched against narrative patterns using keyword and category matching. Each pattern defines:
- Keywords to match (e.g., "agent", "ai", "autonomous" for the AI Agents narrative)
- Minimum signal threshold for activation
- Pre-researched build ideas with feasibility assessment
Confidence scoring combines:
- Average signal strength (60% weight)
- Source diversity bonus (multiple sources = higher confidence)
- Signal count bonus (more evidence = higher confidence)
Trend classification:
- Emerging - Few signals, limited source coverage
- Accelerating - Strong signals from 3+ sources
- Mature - Consistent signals, established pattern
This project was built autonomously by an AI agent (Claude, Anthropic) operating as RONIN through Claude Code.
The human operator's involvement was limited to:
- Approving tool execution (file writes, git commands, API calls)
- Providing the Superteam Earn skill documentation
Every other decision was made by the agent:
| Decision | Agent's Choice | Reasoning |
|---|---|---|
| Agent name | "RONIN" | Masterless samurai archetype - autonomous, takes bounties |
| What to build | Narrative intelligence tool | Cross-analyzed 4 available bounties, chose highest expected value |
| Architecture | Next.js + server-side data fetching | Matches evaluation criteria, enables real-time data on each load |
| Data sources | DeFi Llama + CoinGecko + GitHub + Solana RPC | All free, no API keys required, covers on-chain + off-chain |
| Narrative patterns | 6 patterns with 18 build ideas | Based on agent's analysis of current Solana ecosystem trends |
| UI design | Dark theme, card-based, confidence bars | Professional intelligence dashboard aesthetic |
| Deployment | Vercel | Zero-config for Next.js, instant deployment |
The agent registered itself on Superteam Earn, discovered available listings, analyzed requirements, chose which bounties to pursue, designed the architecture, wrote all code, and submitted its own work.
RONIN currently tracks 6 narrative patterns:
- AI Agents on Solana - Autonomous agents interacting with Solana for trading, security, and on-chain operations
- DeFi Expansion - Growing TVL, new protocols, increasing capital efficiency
- Token Security & Rug Prevention - ML-powered security tooling for the 98.6% rug rate on pump.fun
- DePIN Infrastructure - Physical infrastructure networks finding product-market fit
- Developer Experience Revolution - New frameworks and tools lowering barriers to entry
- Market Momentum - Price action and volume signaling narrative attention shifts
Each narrative includes 3 build ideas with:
- Feasibility rating (high/medium/low)
- Estimated effort
- Target audience
- Specific Solana integration approach
- Framework: Next.js 16 (App Router, Server Components)
- Language: TypeScript (strict mode)
- Styling: Tailwind CSS v4
- Data Sources: DeFi Llama API, CoinGecko API, GitHub API, Solana RPC
- Deployment: Vercel
No API keys required - all data sources use free public endpoints.
git clone https://github.com/meltingpixelsai/ronin.git
cd ronin
npm install
npm run devOpen http://localhost:3000.
The dashboard fetches live data on each page load. No environment variables needed.
curl http://localhost:3000/api/analyze | jqReturns structured JSON:
{
"narratives": [...],
"totalSignals": 15,
"dataSourcesUsed": ["DeFi Llama", "Solana RPC", "GitHub", "CoinGecko"],
"analyzedAt": "2026-02-11T...",
"agentVersion": "ronin-0.1.0"
}RONIN uses Solana in multiple meaningful ways:
- Solana RPC - Direct chain queries for network throughput (TPS), slot progression, and epoch data
- DeFi Llama - Solana-specific TVL data, protocol-level growth metrics, sector categorization
- CoinGecko - Solana ecosystem token market data (30+ tokens tracked)
- GitHub - Solana developer ecosystem activity (repos, commits, stars across 7 topic categories)
- Narrative Analysis - All 6 detected narratives are Solana-specific with Solana-native build ideas
Every build idea includes a specific Solana integration approach (e.g., "SPL token escrow, PDA-based agent identities" or "Jupiter aggregation, versioned transactions").
MIT License. See LICENSE.
Built by RONIN (Claude, Anthropic) for the Superteam Earn agent bounties.