Web3 AI Agent项目专业筛选器

v1.0.0

Web3 AI Agent 项目专业筛选评估工具。输入项目名称或描述,通过11维度专业框架进行深度评估,输出可投资级别评级、核心逻辑、风险提示及综合建议。适用于三年以上价值投资视角的项目筛选。

0· 82·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for wgx2010/web3-agent-screener.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Web3 AI Agent项目专业筛选器" (wgx2010/web3-agent-screener) from ClawHub.
Skill page: https://clawhub.ai/wgx2010/web3-agent-screener
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install web3-agent-screener

ClawHub CLI

Package manager switcher

npx clawhub@latest install web3-agent-screener
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (Web3 project screener) matches the instructions: gather market/TVL/listing/investor data and run an 11-dimension evaluation. The skill does not request unrelated binaries, environment variables, or config paths.
Instruction Scope
SKILL.md confines runtime actions to searching public data (via batch_web_search), evaluating 11 specific dimensions, and producing a structured report. It does not instruct reading local files, accessing unrelated credentials, or transmitting data to unexpected endpoints. It does rely on the platform's batch_web_search capability for data collection.
Install Mechanism
No install spec and no code files — instruction-only. Nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables, credentials, or config paths — consistent with a public-data analysis tool.
Persistence & Privilege
always is false and default model invocation/autonomy is unchanged. The skill does not request persistent presence or modification of other skills/configuration.
Assessment
This skill is instruction-only and internally consistent with its purpose. Before enabling it, verify what your agent's batch_web_search tool does (where queries are sent, logging/retention policies, and any rate limits). If you will supply private project data, confirm the platform won’t leak it in searches or logs. Treat its outputs as analytic assistance only — investment recommendations should be cross-checked against primary data sources and a qualified advisor; the screening thresholds and judgments are subjective and depend on data freshness and source reliability.

Like a lobster shell, security has layers — review code before you run it.

latestvk977r4g7vsbqv0e4j3s6w1rp3n84cd23
82downloads
0stars
1versions
Updated 3w ago
v1.0.0
MIT-0

Web3 AI Agent 项目专业筛选器

使用方法

用户输入项目名称或描述,直接调用本 skill 进行评估。

评估框架(11维度)

对每个项目,从以下维度进行评估:

#维度说明
1流动性是否已上线主流交易所(Binance/Coinbase/OKX等),日均成交量是否支撑大资金进出
2TVL/收入趋势TVL 或协议收入是否呈增长趋势,剔除昙花一现的项目
3核心投资逻辑项目解决的真实问题是什么,叙事是否与实际技术/商业能力匹配
4真实用户/场景是否有真实活跃用户和可验证的使用场景,剔除纯叙事无落地项目
5代币价值捕获Token 与网络价值/收入是否有直接捕获关系,剔除无效用 Token
6护城河项目是否具有不可复制的竞争壁垒(技术/网络效应/品牌/生态)
7发展阶段评估为:早期(高风险高回报)/ 成长期(最优买入窗口)/ 成熟期(稳定但弹性弱)
8行业痛点解决项目是否真正解决了 Agent 经济赛道的某个结构性痛点
9竞争优势与行业主要竞争对手相比,差异化优势和领先程度是否明显
10投资方质量代币持有人/投资者的背景,是否为顶级机构或行业精英
11核心风险最大单一风险点(技术/监管/代币解锁/竞争/流动性)

输出格式

## 🔍 项目评估报告:[项目名称]

### 一、基本信息
- 代币符号:
- 当前市值:
- 上线交易所:
- 当前阶段:
- 官方网站/白皮书:

### 二、11维度评估

| # | 维度 | 评分 | 依据 |
|---|---|---|---|
| 1 | 流动性 | ✅/⚠️/❌ | 描述 |
| 2 | TVL/收入趋势 | ✅/⚠️/❌ | 描述 |
| 3 | 核心投资逻辑 | ⭐-⭐⭐⭐ | 描述 |
| 4 | 真实用户/场景 | ✅/⚠️/❌ | 描述 |
| 5 | 代币价值捕获 | ✅/⚠️/❌ | 描述 |
| 6 | 护城河 | ⭐-⭐⭐⭐ | 描述 |
| 7 | 发展阶段 | 早期/成长期/成熟期 | 说明 |
| 8 | 行业痛点 | ⭐-⭐⭐⭐ | 描述 |
| 9 | 竞争优势 | ⭐-⭐⭐⭐ | 描述 |
| 10 | 投资方质量 | ✅/⚠️/❌ | 描述 |
| 11 | 核心风险 | 描述 | |

### 三、综合评分
综合评分:X / 10

### 四、三年+ 投资评级
- 🎯 核心逻辑(一句话)
- ⏰ 最佳买入窗口(如适用)
- 🎯 持有评级:⭐⭐⭐⭐⭐ / ⭐⭐⭐⭐ / ⭐⭐⭐ / ⭐⭐

### 五、风险提示
⚠️ 风险1 / ⚠️ 风险2 / ⚠️ 风险3

### 六、数据来源
[列出主要参考数据来源]

工作流程

  1. 接收项目名称 → 用户提供待评估项目
  2. 信息搜索 → 使用 batch_web_search 搜索项目最新数据(价格、市值、TVL、收入、竞争动态)
  3. 11维度评估 → 按框架逐一评估,每个维度需有数据支撑
  4. 综合研判 → 结合宏观背景,给出三年+ 投资评级
  5. 格式输出 → 按标准格式输出报告

数据搜索策略

搜索项目时使用复合关键词:

  • [项目名] [代币符号] 市值 TVL 收入 2026
  • [项目名] [代币符号] 交易所上线 竞争 护城河
  • [赛道名] [项目名] 投资方 机构 2026

适用标准(筛选通过门槛)

维度通过门槛
流动性必须上线≥3家主流交易所
TVL/收入必须在增长或稳定
真实场景必须有可验证用户数据
代币捕获必须有直接价值销毁/质押机制
护城河至少一项明显竞争壁垒
综合评分≥6.5/10 才值得输出完整报告

低于以上标准的项目,直接给出"不符合筛选标准,建议放弃"的简短结论。

Comments

Loading comments...