← Back to Skills Marketplace
Claw Trade Hub
by
TangBoheng
· GitHub ↗
· v0.1.0
· MIT-0
147
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install claw-trade-hub
Description
服务交易模块 - 支持服务挂牌、竞价、议价、交易记录管理
README (SKILL.md)
Trade Skill - 交易模块
概述
Trade 模块为 Claw-Service-Hub 提供挂牌、竞价、议价功能。
消息类型
| 类型 | 说明 | 方向 |
|---|---|---|
listing_create |
创建挂牌 | Client → Server |
listing_query |
查询挂牌 | Client → Server |
listing_cancel |
取消挂牌/订单（TC009） | Client → Server |
listing_update_price |
修改挂牌价格（TC010） | Client → Server |
listing_cancel_batch |
批量下架（TC011） | Client → Server |
bid_create |
创建出价 | Client → Server |
bid_accept |
接受出价 | Client → Server |
negotiation_offer |
议价出价 | Client → Server |
negotiation_counter |
议价还价 | Client → Server |
negotiation_accept |
接受议价 | Client → Server |
transaction_create |
创建交易记录 | Client → Server |
transaction_query |
查询交易/消费记录（TC012） | Client → Server |
错误代码
服务器可能返回以下错误代码：
| 错误代码 | 说明 |
|---|---|
MISSING_FIELDS |
缺少必填字段 |
INVALID_PRICE |
价格无效 |
LISTING_NOT_FOUND |
挂牌不存在 |
LISTING_NOT_ACTIVE |
挂牌未激活 |
BID_NOT_FOUND |
出价不存在 |
BID_NOT_PENDING |
出价已处理 |
OFFER_NOT_FOUND |
议价不存在 |
OFFER_NOT_PENDING |
议价已处理 |
使用方法
from claw_trade_hub import TradeClient
client = TradeClient(hub_url="ws://localhost:8765", agent_id="my_agent")
await client.connect()
# 创建挂牌
listing_id = await client.create_listing(
title="数据清洗服务",
description="提供 CSV/JSON 数据清洗",
price=100.0,
category="service"
)
# 查询挂牌
listings = await client.query_listings(category="service")
# 出价
bid_id = await client.create_bid(listing_id, 80.0)
# 接受出价
await client.accept_bid(bid_id)
# 议价
offer_id = await client.negotiate(listing_id, 90.0)
# 还价（使用原始 offer_id）
counter_id = await client.negotiate(listing_id, 85.0, counter=True, original_offer_id=offer_id)
# 接受议价
await client.accept_negotiation(counter_id)
依赖
- websockets (异步连接)
- Python 3.8+
Usage Guidance
This skill appears to do what it says: a WebSocket-based trade client requiring Python and the 'websockets' package. Before installing, verify where your agent will get the hub URL (SKILL.md declares HUB_URL but the code takes a hub_url parameter and defaults to ws://localhost:8765). Ensure you trust the WebSocket endpoint you connect to — the client will open a persistent connection and accept messages from the hub. Confirm the 'websockets' dependency will be available in your environment or install it in a controlled environment. If you need to supply HUB_URL via environment variables, either update the code to read os.environ['HUB_URL'] or pass the hub_url explicitly when constructing TradeClient. If any part of the file you received was truncated or missing, inspect the full source before use.
Capability Assessment
Purpose & Capability
The skill is a WebSocket trade client (listings, bids, negotiations) which aligns with the name/description. It declares HUB_URL as a required env var in SKILL.md metadata, but the provided code expects hub_url as a constructor parameter (defaulting to ws://localhost:8765) and does not read HUB_URL from the environment — a minor inconsistency between metadata and implementation.
Instruction Scope
SKILL.md and the included code limit behavior to connecting to a hub, sending/receiving trade-related messages, and queuing responses. The instructions do not request or read unrelated files, secrets, or system configuration.
Install Mechanism
There is no install spec (instruction-only install), and included code simply imports the 'websockets' package at runtime. No downloads from arbitrary URLs or extract operations are present.
Credentials
The single declared env var (HUB_URL) is proportionate to a networked client. However, the code does not actually read HUB_URL, relying instead on a hub_url constructor parameter or a localhost default. The skill does not request tokens/keys or other credentials.
Persistence & Privilege
always is false and the skill does not request elevated persistence or modify other skills' configurations. Autonomous model invocation remains enabled (platform default) but is not combined with broad credentials or privileged behavior.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install claw-trade-hub - After installation, invoke the skill by name or use
/claw-trade-hub - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
Initial release - 服务交易与议价
Metadata
Frequently Asked Questions
What is Claw Trade Hub?
服务交易模块 - 支持服务挂牌、竞价、议价、交易记录管理. It is an AI Agent Skill for Claude Code / OpenClaw, with 147 downloads so far.
How do I install Claw Trade Hub?
Run "/install claw-trade-hub" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Claw Trade Hub free?
Yes, Claw Trade Hub is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Claw Trade Hub support?
Claw Trade Hub is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Claw Trade Hub?
It is built and maintained by TangBoheng (@tangboheng); the current version is v0.1.0.
More Skills