← 返回 Skills 市场
866
总下载
0
收藏
2
当前安装
1
版本数
在 OpenClaw 中安装
/install research-and-trade
功能描述
Research a token and execute a trade if it passes due diligence. Autonomous research-to-trade pipeline: researches the token, evaluates risk, and only trades if the risk assessment approves. Stops and reports if risk is too high. Use when user wants "research X and buy if it looks good" or "due diligence then trade."
安全使用建议
Do not install or use this skill with real funds until the following are clarified: (1) exactly how and where the skill gets signing authority — which env vars, connectors, or platform wallets it will use; (2) whether every trade requires explicit, per-trade user confirmation and manual signing; (3) what the Task(subagent_type:trade-executor) is and what privileges it has; and (4) what external endpoints (RPC nodes, relayers, third-party services) the skill will call. If you need the research functionality only, prefer using a read-only skill or refusal to execute trades automatically. If you decide to test, use a throwaway wallet with minimal funds on a testnet or small amount and review all prompts and logs. Ask the maintainer for a clear security model and an audit of the trade-executor behavior before trusting it with real assets.
功能分析
Type: OpenClaw Skill
Name: research-and-trade
Version: 0.1.0
The skill is classified as suspicious due to significant prompt injection vulnerabilities. The primary concern is the 'compound context' mechanism described in SKILL.md, where the full, unvalidated outputs from sub-agents (e.g., token-analyst, pool-researcher) are directly embedded into the prompts of subsequent sub-agents (e.g., risk-assessor, trade-executor). This creates a critical vulnerability where a malicious user could craft inputs to an earlier sub-agent, causing it to generate an output containing harmful instructions, which would then be executed by a later, potentially more privileged, sub-agent. This could bypass the skill's intended safety gates and lead to unauthorized actions or data exfiltration, despite the presence of explicit user confirmation steps and hard vetoes for high-risk trades.
能力评估
Purpose & Capability
The skill's stated purpose is an end-to-end research-to-trade pipeline that executes on-chain swaps. However, the package declares no required environment variables, config paths, binaries, or credentials (no private key, wallet connector, RPC URL, or third-party API tokens). A trade-executor subagent is listed but the SKILL.md does not explain how transaction signing or provider access is obtained. This mismatch between claimed capability (executing on-chain trades) and declared requirements is a substantive coherence gap.
Instruction Scope
The runtime instructions let the pipeline automatically proceed from research -> risk assessment -> trade execution (APPROVE → Proceed to Step 4). While the skill mentions vetoes and conditional approvals, it also describes automatically executing trades when approved. The SKILL.md does not describe an explicit user confirmation/signing step, nor does it describe what the trade-executor tool needs or how it obtains consent to spend funds. That grants broad operational discretion to the agent without documenting where signing authority comes from.
Install Mechanism
This is instruction-only with no install spec and no code files — the lowest-risk install surface. There is no download, no package installations, and nothing is written to disk by the skill itself.
Credentials
A trade-capable skill would reasonably require credentials or connectors (wallet private key / wallet provider, RPC endpoints, or an exchange API key). This skill declares none. That absence is suspicious: either the skill assumes hidden platform-level connectors (not documented) or it will attempt to operate without explicit signing or will prompt interactively. Either case should be clarified before use because it affects who can sign and pay gas.
Persistence & Privilege
The skill is not always-enabled (always: false). Autonomous invocation is permitted (disable-model-invocation: false), which is platform-default. Autonomous ability combined with the skill's execution capability is a risk — it could place transactions without clear documented human-in-the-loop signing — but autonomous invocation by itself is not unusual for skills.
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 research-and-trade - 安装完成后，直接呼叫该 Skill 的名称或使用
/research-and-trade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
v0.1.0
Initial release: Autonomous research-to-trade workflow
- Researches a token, analyzes liquidity, evaluates risk, and only executes a trade if approved by due diligence.
- Fully automates and links four expert steps: research, pool selection, risk assessment, and execution.
- Stops immediately if risk is too high, providing a detailed report to the user.
- Gathers and compounds context at each step, enabling richer risk analysis than standalone tools.
- Requires a specific trade amount from the user before proceeding; never guesses trade size.
- Shows progressive findings at each pipeline stage, not just at the end.
元数据
常见问题
Uniswap Research And Trade 是什么？
Research a token and execute a trade if it passes due diligence. Autonomous research-to-trade pipeline: researches the token, evaluates risk, and only trades if the risk assessment approves. Stops and reports if risk is too high. Use when user wants "research X and buy if it looks good" or "due diligence then trade.". 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 866 次。
如何安装 Uniswap Research And Trade？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install research-and-trade」即可一键安装，无需额外配置。
Uniswap Research And Trade 是免费的吗？
是的，Uniswap Research And Trade 完全免费（开源免费），可自由下载、安装和使用。
Uniswap Research And Trade 支持哪些平台？
Uniswap Research And Trade 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 Uniswap Research And Trade？
由 wpank（@wpank）开发并维护，当前版本 v0.1.0。
推荐 Skills