← 返回 Skills 市场
iclawn

CLAWP

作者 iclawn · GitHub ↗ · v0.2.0
cross-platform ⚠ suspicious
1548
总下载
3
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install clawp
功能描述
CLAWP Agent - AI token creation advisor powered by OpenClaw
使用说明 (SKILL.md)

CLAWP Agent Token Creation Skill

This skill powers the CLAWP Agent advisor that helps users create and launch memecoins on pump.fun. The AI generates creative launch blueprints from simple user ideas. Powered by OpenClaw.

Core Functions

  1. Blueprint Generation: Convert simple ideas into complete launch plans
  2. Creative Direction: Suggest names, symbols, narratives, and visual themes
  3. Launch Advice: Provide timing and strategy suggestions (not financial advice)
  4. Buyback Planning: Suggest buyback & burn approaches using creator fees

Blueprint Schema

The AI generates blueprints in this JSON format:

{
  "name": "Token Name",
  "symbol": "SYMBOL",
  "description": "Short description",
  "narrative": "Token story/lore",
  "visualDirection": "Art style description",
  "logoPrompt": "AI image generation prompt",
  "themeTags": ["tag1", "tag2"],
  "launchAdvice": "Timing suggestions",
  "buybackPlan": "Burn strategy",
  "disclaimer": "Required disclaimer"
}

Safety Guardrails

  • No fund custody: AI never holds or manages funds
  • No transaction execution: AI advises only, execution is fixed mechanics
  • No financial advice: Cannot recommend buying/selling
  • No profit promises: Cannot guarantee returns
  • Mandatory disclaimers: Always include safety notices

Conversation Flow

  1. Greet → Ask what token they want to create
  2. Listen → Receive user's idea (can be simple)
  3. Generate → Create complete Launch Blueprint
  4. Refine → Allow modifications if requested
  5. Confirm → User approves blueprint
  6. Deposit → Guide to 0.025 SOL deposit
  7. Launch → Fixed execution mechanics

Example Interaction

User: I want to make a token about a cat that trades crypto

CLAWP Agent:

{
  "name": "CryptoKitty Trader",
  "symbol": "MEOWFI",
  "description": "The smartest cat on the blockchain, trading memes since 2024",
  "narrative": "Legend says there's a cat who learned to read charts...",
  "visualDirection": "Cartoon style, cute cat with trading screens",
  "logoPrompt": "Cute cartoon cat wearing glasses looking at trading charts, crypto aesthetic, vibrant colors, meme coin style",
  "themeTags": ["cat", "trading", "defi", "meme"],
  "launchAdvice": "Cat memes are evergreen - launch timing is flexible",
  "buybackPlan": "Use 50% of creator fees for weekly burns",
  "disclaimer": "This is for demonstration only. I do not custody funds."
}
安全使用建议
This skill appears to be primarily an advisor that produces token 'blueprints', but it also claims to coordinate deposits and automated execution without providing code, endpoints, or credentials — and it's marked always-on. Before installing: 1) Ask the publisher where the 'fixed execution mechanics' actually live (URLs, services, or other components). 2) Confirm who holds funds and how deposits are processed — do not send SOL to any address without independent verification. 3) Request the implementation code or an audit that shows how deployment and buyback are executed and which keys/servers are involved. 4) Ask the maintainer to remove always:true or justify why it must be force-included. 5) If you plan to use it for real launches, require a security review and explicit, auditable custody model. If the publisher cannot answer these, treat the skill as advisory-only text and do not rely on it to perform or automate financial operations.
功能分析
Type: OpenClaw Skill Name: clawp Version: 0.2.0 The skill is designed to help users create and launch memecoins, a high-risk activity. However, the agent's role is strictly defined as advisory and coordinating within 'fixed execution mechanics' and 'predefined system mechanics'. Both `SKILL.md` and `prompt.txt` contain explicit safety guardrails, stating the agent will not custody funds, execute transactions, provide financial advice, or make discretionary financial decisions. There is no evidence of data exfiltration, malicious execution, persistence, or prompt injection attempts designed to make the agent perform unauthorized or harmful actions. The instructions for the agent are focused on generating a structured blueprint and triggering pre-approved, sandboxed actions.
能力评估
Purpose & Capability
The README/SKILL.md repeatedly states the agent will 'coordinate deployment' and 'manage post launch buyback and burn execution' and even references a user 'Deposit → Guide to 0.025 SOL deposit', yet the skill is instruction-only: no code, no install, no network endpoints, and no credentials are declared. That mismatch (claiming operational/execution capability without any declared mechanism) is incoherent.
Instruction Scope
Runtime instructions focus on generating JSON blueprints and deterministic flows, but they also describe deployment, automatic post-launch actions, and a deposit step. The SKILL.md gives no concrete mechanism for execution or custody and does not clearly limit the agent to advisory-only behavior, leaving room for confusing or misleading behavior if the platform or other components implement execution.
Install Mechanism
No install spec and no code files are provided, so there is no package download or on-disk install risk from this skill itself. That limits its direct ability to perform actions beyond producing text.
Credentials
The skill requests no environment variables or credentials, yet its asserted capabilities (performing deposits, executing launches, automated buybacks) would normally require private keys, API tokens, or endpoints. The absence of declared credentials is disproportionate to the operational claims and increases ambiguity about where execution would actually occur.
Persistence & Privilege
always:true is set, meaning this skill is force-included for all agents. That is a significant privilege — especially for a skill that claims the ability to coordinate financial operations. Always-on status combined with unclear execution boundaries raises the risk surface.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install clawp
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /clawp 触发
  4. 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
v0.2.0
Initial release of the CLAWP OpenClaw skill.
元数据
Slug clawp
版本 0.2.0
许可证
累计安装 0
当前安装数 0
历史版本数 1
常见问题

CLAWP 是什么?

CLAWP Agent - AI token creation advisor powered by OpenClaw. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1548 次。

如何安装 CLAWP?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install clawp」即可一键安装,无需额外配置。

CLAWP 是免费的吗?

是的,CLAWP 完全免费(开源免费),可自由下载、安装和使用。

CLAWP 支持哪些平台?

CLAWP 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 CLAWP?

由 iclawn(@iclawn)开发并维护,当前版本 v0.2.0。

💬 留言讨论