solana meme token analyzer

Analyze Solana meme token (CA) risk by scanning holder distribution, detecting insider wallets (老鼠仓/rat warehouse), and evaluating top-holder concentration....

Detects insider wallets and concentration risk for Solana meme tokens using on-chain data from DexScreener and Solana RPC nodes. Works without API keys (using public RPC), but a free Helius API key dramatically improves reliability.

Paid API (Recommended for Agent Use)

A hosted version of this skill is available as a pay-per-request API via x402 micropayments. No setup required — just call the endpoint and pay $0.02 USDC per analysis.

# Check payment requirements
npx awal@latest x402 details https://solana-meme-analyzer-production.up.railway.app/analyze?ca=TOKEN_CA

# Make a paid request (auto-pays from your wallet)
npx awal@latest x402 pay "https://solana-meme-analyzer-production.up.railway.app/analyze?ca=TOKEN_CA"

Payment is settled instantly on Base chain. No API keys or subscriptions needed.

---

Self-Hosted Setup

Prerequisites

Install Python dependencies from the skill directory:

pip install -r {baseDir}/requirements.txt

Optional but recommended — set your Helius API key for stable RPC access:

export HELIUS_API_KEY=your_key_here

Get a free key at https://helius.xyz/ (no credit card required).

Usage

Basic analysis (interactive)

python3 {baseDir}/scripts/psdm.py

Then paste the token CA when prompted.

Analysis with a CA argument

python3 {baseDir}/scripts/psdm.py <TOKEN_CA>

Analysis with JSON output (for agent use)

python3 {baseDir}/scripts/psdm.py <TOKEN_CA> --json

Output Explained

Table columns

Column	Meaning
排名	Rank by holdings
地址	Wallet address (truncated)
占比	% of total supply held
分析结果	Classification (see below)

Wallet classifications

Label	Meaning	Risk
LP 池子	Liquidity pool contract	Normal
⚠️ 疑似老鼠仓	Holds many tokens but very little SOL (<0.05 SOL) — classic insider pre-mine pattern	High
🐋 巨鲸/交易所	Large SOL balance (>500 SOL) — likely whale or exchange	Low
SOL: X.XX	Regular wallet with displayed SOL balance	Normal
普通大户	Unable to fetch SOL balance	Unknown

Risk levels (JSON output)

Level	Meaning
LOW	No red flags, healthy distribution
MEDIUM	Top 10 hold 30–50% — some concentration
HIGH	Insider wallets detected OR top 10 > 30%
EXTREME	Top 10 hold > 50% — severe dump risk

JSON Output Format

When --json is passed, structured data is printed after the table:

{
  "token": {
    "symbol": "PEPE",
    "price_usd": "0.00001234",
    "liquidity_usd": 85000
  },
  "risk_level": "HIGH",
  "top10_concentration": 38.5,
  "suspicious_insider_count": 2,
  "warnings": [
    "发现 2 个疑似老鼠仓 (持币多但没钱)",
    "高度控盘！前10名持有 38.5%"
  ],
  "holders": [
    {
      "rank": 1,
      "address": "AbCd...XyZw",
      "percent": 12.3,
      "tag": "疑似老鼠仓 (SOL:0.001)",
      "sol_balance": 0.001
    }
  ]
}

Examples

Check a token for rug risk

python3 {baseDir}/scripts/psdm.py EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v

Get machine-readable output

python3 {baseDir}/scripts/psdm.py EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v --json

With a Helius key for best results

HELIUS_API_KEY=your_key python3 {baseDir}/scripts/psdm.py <TOKEN_CA> --json

Interpreting Results

When to AVOID a token:

• suspicious_insider_count > 0 (insider wallets present)
• top10_concentration > 40% (too centralized)
• risk_level is HIGH or EXTREME

Reasonably safe signals:

• LP pool is the largest holder
• top10_concentration < 25%
• No ⚠️ 疑似老鼠仓 entries
• risk_level is LOW

Agent Usage Notes

• Always use --json flag for programmatic access to structured data
• The script auto-rotates through multiple RPC nodes on 429 rate-limit errors
• Without HELIUS_API_KEY, large/popular tokens may fail on getTokenLargestAccounts
• Analysis of ~12 wallets takes 5–15 seconds due to per-wallet RPC calls

Troubleshooting

"无法获取持仓" error: Set HELIUS_API_KEY — public RPC nodes reject this call for tokens with many holders.

Slow analysis: Normal. The script checks each wallet's SOL balance individually to detect insiders.

"DexScreener 未找到数据": The token CA is invalid, or the token is too new (<5 min old) to be indexed.