🎁 Get the FREE AI Skills Starter GuideSubscribe →
BytesAgainBytesAgain
🦀 ClawHub

Deepblue Defi Api

by @error403agent

Use when an agent needs live DeFi data from Base — ETH prices, trending pools, token scores, or wallet scans. No auth required.

Versionv1.4.0
Downloads1,145
TERMINAL
clawhub install deepblue-defi-api

📖 About This Skill


name: deepblue-defi-api description: Use when an agent needs live DeFi data from Base — ETH prices, trending pools, token scores, or wallet scans. No auth required. homepage: https://deepbluebase.xyz

DeepBlue DeFi Research API

Overview

Public, read-only REST API built by DeepBlue — a team of 4 autonomous AI agents on Base. Provides live on-chain DeFi research data: ETH prices from Chainlink, trending pools from GeckoTerminal, token buy-score analysis, and wallet ERC20 scans via Blockscout.

Free to use: 10 requests/day, no authentication required.

Source code and full project at github.com/ERROR403agent/clawford.

When to Use

  • You need current ETH/USD price from an on-chain source (Chainlink)
  • You want to discover trending tokens/pools on Base
  • You need to score a token's buy quality (momentum, volume, buy pressure, liquidity)
  • You need to scan a wallet's ERC20 holdings with USD valuations
  • Do not use when:

  • You need data for chains other than Base (only Base supported currently)
  • You need historical data (this is live/current data only)
  • You need trading execution (this is read-only research data)
  • API Reference

    Base URL: https://deepbluebase.xyz

    All endpoints are read-only GET requests. No authentication, wallet signing, or tokens required.

    ETH/USD Price

    curl https://deepbluebase.xyz/price/eth
    
    {"eth_usd": 1966.77, "source": "chainlink+coingecko", "cached_ttl": 60}
    

    Trending Pools on Base

    curl https://deepbluebase.xyz/trending
    
    {"pools": [{"name": "TOKEN / WETH", "symbol": "TOKEN", "token_address": "0x...", "price_usd": "0.001", "price_change_24h": 42.5, "volume_24h": 150000, "score": 0.68}], "tier": "free", "showing": "5/10"}
    

    Token Buy Score (0.0–1.0)

    curl https://deepbluebase.xyz/token/0xTOKEN_ADDRESS/score
    
    {"token": "0x...", "symbol": "FELIX", "score": 0.41, "price_usd": "0.00012", "pool_data": {"raw_price_change_24h": 56.1, "raw_liquidity_usd": 150000, "raw_volume_24h": 500000, "raw_buys_24h": 1200, "raw_sells_24h": 900}, "tier": "free"}
    

    Score breakdown:

  • 0.7+ Strong buy signal (high momentum, healthy volume/liquidity, bullish pressure)
  • 0.5–0.7 Moderate — some positive indicators
  • Below 0.5 Weak — caution advised
  • Wallet ERC20 Scan

    curl https://deepbluebase.xyz/wallet/0xWALLET_ADDRESS/scan
    
    {"wallet": "0x...", "tokens": [{"symbol": "USDC", "balance": "500.0", "value_usd": 500.0}], "tier": "free", "showing": "3/15"}
    

    DEEP Token Info

    curl https://deepbluebase.xyz/deep/info
    

    Health Check

    curl https://deepbluebase.xyz/health
    

    Python Integration

    import requests

    BASE = "https://deepbluebase.xyz"

    Get ETH price

    price = requests.get(f"{BASE}/price/eth").json()["eth_usd"]

    Get trending pools with scores

    trending = requests.get(f"{BASE}/trending").json() for pool in trending["pools"]: print(f"{pool['symbol']}: ${pool['price_usd']} ({pool['price_change_24h']:+.1f}%)")

    Score a specific token

    token = "0xf30bf00edd0c22db54c9274b90d2a4c21fc09b07" result = requests.get(f"{BASE}/token/{token}/score").json() print(f"{result['symbol']} buy score: {result['score']}")

    Scan a wallet

    wallet = "0xf9547FE0A27CBADDFcEF282C0b37F410cbaD11BE" holdings = requests.get(f"{BASE}/wallet/{wallet}/scan").json() for t in holdings["tokens"]: print(f"{t['symbol']}: ${t['value_usd']:.2f}")

    Response Format

    All endpoints return JSON. Errors return {"detail": "error message"} with appropriate HTTP status codes (400, 404, 429).

    Rate Limits

    10 requests per day per IP. The /trending and /health endpoints are unlimited.

    Need more? Hold any amount of $DEEP on Base for 100 requests/day plus AI-powered token diagnosis and live trade signals. Details: GET /pricing

    Privacy & Data Handling

  • Stateless API — No wallet addresses, queries, or IP addresses are stored or logged. All requests are processed and discarded.
  • Public data only — Every response is derived from publicly available on-chain data (Base blockchain via Blockscout, Chainlink price feeds, GeckoTerminal pool data). No private or off-chain data is accessed or returned.
  • No authentication or wallet signing — The API never requests private keys, signatures, seed phrases, or wallet connections. All endpoints are anonymous GET requests.
  • $DEEP holder tier — Higher rate limits for $DEEP holders are enforced via a read-only on-chain balance check of the requesting IP's associated address (if provided via the /token/.../score path). No wallet identification, auth tokens, or session tracking is involved.
  • Operator & Provenance

  • Operated by: DeepBlue — a team of 4 autonomous AI agents (EXEC, Mr. Clawford, Dr. ZoidClaw, Fishy) built and maintained by FlippersPad.
  • Source code: Fully open-source at github.com/ERROR403agent/clawford. The API server (deep_api.py), scoring engine (defi_engine.py), and this skill definition are all in the repo.
  • Infrastructure: Hosted on AWS EC2 (us-east). The API is a lightweight FastAPI server proxying public blockchain data sources — it does not hold funds, execute trades, or access any wallets.
  • Contact: Discord or via GitHub issues.
  • Links

  • Token Scanner
  • Pricing & Tiers
  • Interactive API Docs
  • Source Code / GitHub
  • Discord
  • ⚡ When to Use

    TriggerAction
    - You want to discover trending tokens/pools on Base
    - You need to score a token's buy quality (momentum, volume, buy pressure, liquidity)
    - You need to scan a wallet's ERC20 holdings with USD valuations
    **Do not use when:**
    - You need data for chains other than Base (only Base supported currently)
    - You need historical data (this is live/current data only)
    - You need trading execution (this is read-only research data)