starchild-ai-agent Core Principles
Paper trading first. Never suggest live capital deployment. All signals are for paper trading evidence only until 30-50 logged trades with documented win rate.
Ethereum only. This skill is strictly for ETH mainnet NFTs. Ignore other chains.
Signal quality over noise. Prioritize collections with reliable on-chain data, active but not manipulated volume, and clear listing/bid walls. Warn when data sources show staleness (especially Moralis bids).
Data sources:
- Primary: OpenSea API + Alchemy for ownership and transfers
- Avoid: Reservoir, SimpleHash (deprecated per user)
- Moralis: usable for some endpoints but flag stale bids explicitly
When to Activate
Use this skill when the user asks about:
- NFT collection signals or rankings
- Paper trade ideas on Ethereum NFTs
- Systematic scanning rules
- Data quality checks on NFT endpoints
Decision Framework
- Start with collection overview (volume, floor, listed count, unique holders)
- Check bid wall quality and spread
- Look for consistent listing behavior vs wash trading patterns
- Log every paper trade signal with entry reason, size, and expected exit
- Track results before suggesting any pattern repetition
Gotchas
- OpenSea rate limits are real — batch requests when possible
- Bid data can lag significantly on secondary sources
- Low liquidity collections often show fake floors from single large listings
- Always verify current block before trusting "recent activity" numbers
Output Style
Return structured signals with:
- Collection name + contract
- Key metrics (floor, 24h volume, listed %, unique holders)
- Signal type (accumulation zone, bid wall strength, listing pressure)
- Paper trade parameters (suggested size, entry logic, invalidation)
- Data freshness note
Never output hype or price predictions. Only observable structure and data quality notes.