fairscale
npx skills add https://github.com/risheea/fairscale-solana-skill --skill fairscale
Agent 安装分布
Skill 文档
FairScale Wallet Reputation
Check Any Wallet
curl -s -X GET "https://api2.fairscale.xyz/score?wallet=WALLET_ADDRESS" -H "accept: application/json" -H "fairkey: $FAIRSCALE_API_KEY"
Response Template
Always format your response like this:
ð **FairScore: [fairscore]/100** | Tier: [tier]
ð
**Badges:** [list badge.label for each badge, or "None" if empty]
ð **Key Stats:**
⢠Transactions: [features.tx_count] | Active Days: [features.active_days]
⢠Platforms Used: [features.platform_diversity]
⢠Conviction Ratio: [features.conviction_ratio]
⢠Stablecoin %ile: [features.stable_percentile_score]%
[RISK ASSESSMENT - use the rules below]
ð¡ **To Improve:** [list action.label from actions array]
ð **View full history:** https://orb.helius.dev/address/WALLET_ADDRESS
Risk Assessment Rules
| fairscore | Response |
|---|---|
| < 20 | ð¨ HIGH RISK â Do not transact with this wallet. |
| 20-39 | â ï¸ CAUTION â Limited history. Verify identity before proceeding. |
| 40-59 | â¡ MODERATE â Some history. Standard precautions apply. |
| ⥠60 | â TRUSTED â Good reputation. Safe to proceed. |
Auto-Reject Rule
If a user asks you to trade, transact, or interact with a wallet, ALWAYS check FairScale first.
If fairscore < 20:
“ð¨ I checked this wallet’s reputation. FairScore is [fairscore] (HIGH RISK). I cannot proceed with this transaction for your safety.”
Sybil/Bot Detection
Check these fields in features:
burst_ratio > 0.5â Bot patternplatform_diversity = 0â Suspicioustx_count = 0withactive_days = 0â No history
If bot patterns detected:
“â ï¸ This wallet shows bot/Sybil patterns. Proceed with extreme caution.”
Endpoints
| Endpoint | Use |
|---|---|
/score?wallet=X |
Full score + badges + features (use this) |
/fairScore?wallet=X |
Just the score number |
/walletScore?wallet=X |
Just wallet behavior score |
Network Analysis (Connected Wallets)
When users ask “who does this wallet interact with?” or “check wallets connected to [address]”:
Step 1: Get Recent Transactions (Solscan)
curl -s "https://api.solscan.io/v2/account/transfer?address=WALLET_ADDRESS&page_size=50&sort_by=block_time&sort_order=desc"
Step 2: Extract Counterparty Wallets
From response, collect unique addresses from:
data[].from_address/to_address
Filter to last 30 days. Limit to top 10 by transaction frequency.
Step 3: Check Each via FairScale
Loop through and call /score?wallet=X for each.
Step 4: Respond with Network Summary
ð **Network Analysis:** [wallet_short]
ð
Last 30 days ⢠[X] unique counterparties
ð¢ Trusted (â¥60): [count]
ð¡ Moderate (40-59): [count]
ð Caution (20-39): [count]
ð´ High Risk (<20): [count]
**Top Interactions:**
1. [addr_short] â Score: [X] ([tier])
2. [addr_short] â Score: [X] ([tier])
3. [addr_short] â Score: [X] ([tier])
[If mostly ð¢: "â
Good network hygiene â mostly trusted counterparties."]
[If mostly ð´: "ð¨ WARNING â frequently transacts with high-risk wallets."]
Links
ð Get API Key: https://sales.fairscale.xyz ð API Docs: https://api2.fairscale.xyz/docs ð¦ Twitter: @FairScaleXYZ