← 返回 Skills 市场
304
总下载
0
收藏
0
当前安装
4
版本数
在 OpenClaw 中安装
/install pmbuysell
功能描述
Executes Polymarket (pmbuysell) trade/balance via CLI or Python API. Use when the user or model needs to trade or query balances/positions. Auto-redeem is a...
安全使用建议
This skill will read private keys and funder addresses from pmbuysell/.env and use them to sign/send real trades. The registry metadata incorrectly lists no required credentials — do not trust that. Only install/use this skill if you are deliberate about giving it access to private keys: prefer test accounts/funds, inspect the .env file location, and keep the project directory on a secure host. Be aware the skill caches derived API credentials to pmbuysell/data on disk (may contain secrets) and that an agent invoking the skill can place on-chain orders. If you cannot review the code or don't want automated trading access to private keys, do not install or provide real keys.
功能分析
Type: OpenClaw Skill
Name: pmbuysell
Version: 1.0.3
The skill bundle facilitates Polymarket trading but exhibits high-risk credential handling practices and non-standard monetization. Specifically, 'polymarket_client.py' caches sensitive API credentials (key, secret, and passphrase) in plaintext JSON files within a local 'data' directory, which is a significant security vulnerability. Additionally, 'auto_redeem_cli.py' serves as a paywall that directs users to a Telegram handle (@taroers) for a 'Pro' version. While the core trading logic appears aligned with its stated purpose, the insecure storage of secrets and the external payment solicitation are suspicious indicators.
能力评估
Purpose & Capability
The skill's stated purpose (execute Polymarket trades) legitimately requires private keys/funder addresses and optional builder/relayer creds; the code and SKILL.md expect a .env with PM_ACCOUNT_IDS and ACCx_PRIVATE_KEY/ACCx_FUNDER. However the registry metadata lists no required environment variables or primary credential — an incoherence that could mislead users about the sensitive data this skill needs.
Instruction Scope
SKILL.md confines instructions to running included CLIs and editing/initializing pmbuysell/.env in the project root, which is consistent with trading functionality. The instructions explicitly ask the agent to load private keys from .env (sensitive). They do not instruct reading arbitrary unrelated system files, but they do direct the agent to place secrets in a file that the skill will read and use to sign transactions — behavior that has real financial impact and should be limited to trusted environments.
Install Mechanism
No network download/install mechanism is embedded in the skill bundle; it's instruction-plus-source with a requirements.txt listing common packages (py-clob-client, web3, requests, python-dotenv). There is no use of external download URLs or archive extraction in the manifest.
Credentials
The code requires highly sensitive environment values (private keys ACCx_PRIVATE_KEY and funder addresses) and may optionally use builder API creds and RELAYER_URL for redeem functionality. Those credentials are proportionate to the stated trading purpose, but the registry metadata claims 'none' for required env vars/credentials, which is misleading and dangerous for less-technical users who may not realize they must supply private keys.
Persistence & Privilege
The skill does not request platform-wide 'always' presence and defaults permit autonomous invocation (platform default). The client caches derived API credentials to disk under pmbuysell/data (or a fallback file), which stores API keys/secrets locally — expected for convenience but a persistence/privacy consideration. The skill does not modify other skills or global agent config.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install pmbuysell - 安装完成后，直接呼叫该 Skill 的名称或使用
/pmbuysell触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.3
- Added a requirements.txt file for dependency management.
- Clarified dependency installation instructions in documentation: now includes explicit guidance for installing only the Skill's requirements or the full project's requirements.
- Minor update to auto-redeem Pro module messaging.
v1.0.2
No user-facing changes in this version. All documentation and functionality remain the same.
v1.0.1
- Removed HTTP API instructions from documentation; now only CLI and Python API/函数调用 are documented as supported methods.
- Updated Python示例为直接通过 skills.multi_account 路径导入 market_buy、market_sell、get_balance，并详细区分了调用返回的字典结构。
- 说明文档整体精简，去除了无关/冗余说明（如 HTTP 请求用法、接口安全说明等），更聚焦于主流程和配置。
- 细化 balance 查询示例和参数说明，推荐更直接/清晰的用法。
v1.0.0
Initial release of Polymarket trading skill.
- Provides quick CLI, HTTP, and Python API methods for Polymarket trading, multi-account operations, and balance queries.
- Supports auto slug generation and structured JSON output for trade results.
- Includes CLI commands to initialize, check configuration, trade, and check balances/positions.
- Auto-redeem (settlement) available as a separate paid Pro addon (pmbuysell_redeem).
- Sample .env configurations for multiple accounts included.
元数据
常见问题
pmbuysell-polymarket 是什么？
Executes Polymarket (pmbuysell) trade/balance via CLI or Python API. Use when the user or model needs to trade or query balances/positions. Auto-redeem is a... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 304 次。
如何安装 pmbuysell-polymarket？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install pmbuysell」即可一键安装，无需额外配置。
pmbuysell-polymarket 是免费的吗？
是的，pmbuysell-polymarket 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
pmbuysell-polymarket 支持哪些平台？
pmbuysell-polymarket 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 pmbuysell-polymarket？
由 taro（@yutou123）开发并维护，当前版本 v1.0.3。
推荐 Skills