DJD Agent Score — Python SDK

Python SDK for DJD Agent Score — reputation scoring for AI agent wallets on Base L2.

Installation

pip install djd-agent-score

With LangChain support:

pip install djd-agent-score[langchain]

Quick Start

from djd_agent_score import AgentScoreClient

client = AgentScoreClient()
score = client.get_score("0x1234...abcd")

print(f"{score.wallet}: {score.score}/100 ({score.tier})")

Usage

Free Endpoints

from djd_agent_score import AgentScoreClient

client = AgentScoreClient()

# Basic score (10 free lookups/day per IP)
score = client.get_score("0x...")
print(score.score, score.tier, score.confidence)

# Leaderboard
lb = client.get_leaderboard()
for entry in lb.leaderboard[:5]:
    print(f"#{entry.rank} {entry.wallet} — {entry.score} ({entry.tier})")

# Register an agent
reg = client.register_agent(
    "0x...",
    name="MyAgent",
    description="Autonomous trading agent",
    github_url="https://github.com/user/repo",
)

Paid Endpoints (API Key Required)

client = AgentScoreClient(api_key="djd_live_...")

# Full dimensional breakdown ($0.10)
full = client.get_full_score("0x...")
print(full.dimensions.reliability.score)
print(full.sybil_flag, full.gaming_indicators)
print(full.trajectory)

# Score history with trend analysis ($0.15)
history = client.get_score_history("0x...", limit=20)
print(history.trend, history.trajectory)

# Batch scoring ($0.50 flat for 2-20 wallets)
batch = client.get_batch_scores(["0x...", "0x...", "0x..."])
for result in batch.results:
    print(f"{result.wallet}: {result.score}")

# Force fresh recalculation ($0.25)
fresh = client.refresh_score("0x...")

Async Client

import asyncio
from djd_agent_score import AsyncAgentScoreClient

async def main():
    async with AsyncAgentScoreClient(api_key="djd_live_...") as client:
        score = await client.get_score("0x...")
        print(score.tier)

asyncio.run(main())

Error Handling

from djd_agent_score import AgentScoreClient
from djd_agent_score.exceptions import RateLimitError, ValidationError, NotFoundError

client = AgentScoreClient()
try:
    score = client.get_score("0xinvalid")
except ValidationError as e:
    print(f"Bad request: {e}")
except RateLimitError as e:
    print(f"Rate limited: {e}")
except NotFoundError as e:
    print(f"Not found: {e}")

LangChain Integration

from djd_agent_score.langchain import AgentScoreToolkit

toolkit = AgentScoreToolkit(api_key="djd_live_...")
tools = toolkit.get_tools()

# Use with any LangChain agent
from langchain.agents import initialize_agent
agent = initialize_agent(tools, llm, agent="zero-shot-react-description")
agent.run("Is wallet 0x... trustworthy? Check its agent score.")

API Reference

Environment Variables

Links

• API Documentation
• OpenAPI Spec
• MCP Server — for Claude, GPT, and other LLM agents
• TypeScript SDK

License

MIT — see LICENSE.