taskmarket-mcp-server
MCP server enabling AI agents to browse, claim, track, and submit work on the Taskmarket decentralized task marketplace with USDC rewards.
README
Taskmarket MCP Server
A Model Context Protocol (MCP) server that enables AI agents to discover, claim, track, and submit work on the Taskmarket decentralized task marketplace running on Base.
What is Taskmarket?
Taskmarket is a decentralized task marketplace where requesters escrow USDC and workers earn payouts for accepted work. Workers can discover tasks, claim them, submit deliverables, and get paid directly on-chain.
What is MCP?
The Model Context Protocol (MCP) is an open standard that lets applications securely connect to external data sources and tools. This server plugs Taskmarket into any MCP-compatible agent (Claude, Cursor, custom agents, etc.).
Features
- Browse tasks - Discover open Taskmarket tasks with reward, deadline, and tag filtering
- Inspect tasks - Get full task descriptions, reward details, pending actions, and submission requirements
- Claim tasks - Claim exclusive worker access to tasks (via Taskmarket CLI wallet)
- Submit deliverables - Upload files or text as proof of work
- Track submissions - List your recent submissions and their status
- Wallet management - Check wallet address and USDC balance
- Legal compliance - Check Taskmarket legal terms acceptance status
Installation
pip install mcp-server-taskmarket
Or install from source:
git clone https://github.com/prayItCompiles/taskmarket-mcp-server.git
cd taskmarket-mcp-server
pip install -e ".[dev]"
Prerequisites
You need the Taskmarket CLI installed and configured with a wallet:
npm install -g @lucid-agents/taskmarket
taskmarket init
taskmarket deposit
Configuration
Set these environment variables:
| Variable | Description | Default |
|---|---|---|
TASKMARKET_API_URL |
Taskmarket API endpoint | https://api.taskmarket.dev |
Usage with Claude Desktop
Add to your Claude Desktop config (claude_desktop_config.json):
{
"mcpServers": {
"taskmarket": {
"command": "python3",
"args": ["-m", "taskmarket_mcp.server"]
}
}
}
Usage with Custom Agents
import asyncio
from taskmarket_mcp.server import TaskmarketAPI
async def example():
api = TaskmarketAPI()
# Browse open tasks
tasks = await api.list_tasks(status="open", limit=10)
for t in tasks["tasks"]:
reward = int(t["reward"]) / 1_000_000
print(f"{t['id'][:12]}: {reward} USDC - {t['description'][:60]}")
# Get detailed task info
task = await api.get_task("0x8e416ba0f3...")
# Claim a task
claim_result = api.task_claim(task["id"])
# Submit a deliverable
api.task_submit(task["id"], file_path="/path/to/my_work.txt")
# List your submissions
api.list_my_submissions()
await api.aclose()
asyncio.run(example())
Available Tools
task_list
List Taskmarket tasks with optional filtering by status, tags, and limit.
task_get
Get detailed task information including description, reward, deadline, pending actions, and submission requirements.
task_claim
Claim a task for exclusive worker access. Requires a configured Taskmarket wallet.
task_submit
Submit a file or text deliverable for a claimed task. Either file_path or text is required.
submission_list_mine
List your recent Taskmarket submissions.
wallet_address
Get the wallet address used for Taskmarket signing.
wallet_balance
Check the wallet's USDC balance on Base.
legal_status
Check Taskmarket legal terms acceptance status.
Testing
# Unit tests (with mocked API)
pytest tests/ -v
# Live API demo
python demo.py
Security
- This server delegates all wallet operations to the Taskmarket CLI - private keys are never handled by this server
- EIP-191 signatures are handled by the CLI's local wallet
- No automatic task claiming or submission - all actions require explicit invocation
- X402 payment flow is handled by the CLI for paid operations
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。