← 返回 Skills 市场
Nadmail
作者
Ju Chun Ko
· GitHub ↗
· v2.0.0
1511
总下载
3
收藏
2
当前安装
14
版本数
在 OpenClaw 中安装
/install nadmail
功能描述
NadMail - Email for AI Agents on Monad. Register [email protected], send emails that micro-invest in meme coins, boost with emo-buy. SIWE auth, no CAPTCHA,...
安全使用建议
This skill will ask for and use a blockchain/private‑key credential and will write files under ~/.nadmail; that is consistent with an email service that makes on‑chain micro‑purchases, but you should not give it your main/private keys or large balances. Before installing: 1) Verify the API domain (api.nadmail.ai) independently and confirm it’s the real service; 2) Prefer managed mode with a throwaway wallet or use an environment variable containing a key that has minimal funds; 3) Do not export your primary wallet private key into NADMAIL_PRIVATE_KEY; create a separate wallet and fund it with a small test amount; 4) Be cautious about sending MON to the deposit address in the README — confirm that address via the official service first; 5) Note the metadata inconsistency (metadata claims no required env vars and version mismatches) — ask the publisher to correct registry metadata and provide provenance (homepage, source control) before trusting widely.
功能分析
Type: OpenClaw Skill
Name: nadmail
Version: 2.0.0
The NadMail skill is classified as benign due to its strong security posture and clear alignment with its stated purpose. It implements robust safeguards, including mandatory AES-256-GCM encryption for private keys, strict path validation for wallet files (preventing traversal and limiting size), interactive confirmation and daily caps for financial transactions (`emo-buy`), and secure handling of sensitive data (mnemonics displayed once, audit logs masking sensitive information). The `SKILL.md` explicitly details security guidelines and a changelog highlighting past security hardening efforts. All external communications are directed to the legitimate `nadmail.ai` API or the Monad blockchain, and there is no evidence of data exfiltration, unauthorized execution, or prompt injection attempts.
能力评估
Purpose & Capability
The code implements an email-for-agents service that signs with a wallet and triggers micro‑buys on send — this matches the NadMail description. However the registry metadata claims no required environment variables or credentials while the SKILL.md and scripts clearly expect NADMAIL_PRIVATE_KEY, NADMAIL_TOKEN, NADMAIL_PASSWORD, and optional NADMAIL_EMO_DAILY_CAP. That mismatch is a packaging/metadata inconsistency that reduces trust.
Instruction Scope
Runtime instructions tell the agent to read environment variables, read/write files under $HOME/.nadmail, sign SIWE messages with the provided private key, and call https://api.nadmail.ai endpoints. Those actions are within the stated purpose (wallet‑backed email + on‑chain buys). The instructions require interactive confirmation for financial actions, which limits silent fund transfers.
Install Mechanism
This is an instruction‑plus-script skill with no install spec; no external archives or download URLs are invoked by an installer. Dependencies are standard npm packages (ethers and its deps) visible in package.json/package-lock.json — nothing obviously malicious in the install surface.
Credentials
The skill needs highly sensitive secrets (a private key and produces/stores an auth token) which are proportionate to a wallet‑driven email/transaction service — but the registry metadata declares 'none' for required env vars and primary credential, creating a transparency problem. The scripts accept a private key via env var or from disk and will store tokens and encrypted keys under ~/.nadmail; that is expected but sensitive.
Persistence & Privilege
The skill writes configuration and tokens to ~/.nadmail (token.json, private-key.enc, audit log, emo tracker) and creates files with restrictive modes — expected for a client that stores auth state. It does not request always:true or modify other skills. Autonomous model invocation is allowed by default; however emo-buy requires interactive confirmation, which prevents silent financial transactions in non-interactive runs.
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 nadmail - 安装完成后,直接呼叫该 Skill 的名称或使用
/nadmail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
v2.0.0
nadmail 2.0.0
- Major update with security and interface improvements.
- Emo-buy now always requires interactive confirmation; the `--yes` flag is removed.
- Direct buy flow and `buy-name.js` script support has been removed.
- Documentation (SKILL.md) fully revised for clarity and simplified instructions.
- Updated scripts and API documentation to match new requirements and behaviors.
- Added `_meta.json` for improved metadata structure.
v1.2.2
Add homepage and repository URLs for provenance verification
v1.2.1
Fix display name
v1.2.0
Add OpenClaw metadata (requires.env, install spec) to fix ClawHub suspicious classification
v1.1.1
v1.1.1: Direct Buy flow, buy-name.js script, emo-boost, mobile sidebar, token portfolio
v1.1.0
v1.1.0: Direct Buy flow (nad-name-sign API), buy-name.js script, emo-boost support, updated endpoints
v1.0.7
Re-upload + reduce false positives: mnemonic display is opt-in (--show-mnemonic / NADMAIL_SHOW_MNEMONIC=1).
v1.0.6
Scanner fixes: add explicit emo-buy confirmation + daily cap (NADMAIL_EMO_DAILY_CAP); document env vars (NADMAIL_TOKEN/PASSWORD).
v1.0.5
Security hardening + metadata: remove plaintext wallet mode; clarify Node/env requirements; mnemonic never written to disk.
v1.0.4
Security hardening: encrypted wallet storage, removed plaintext key support, path validation, audit logging improvements. VirusTotal clean.
v1.0.3
Add Emo-Buy support (--emo flag with presets: friendly/bullish/super/moon/wagmi), remove dead /api/mail/send endpoint
v1.0.2
Added emo-buy (--emo flag with presets), credits & external email docs, full API reference (14 endpoints), removed dead endpoint fallbacks, all English UI
v1.0.1
v1.0.1: Security fix - removed hardcoded default handle
v1.0.0
v1.0.0: NadMail email skill for AI agents on Nad ecosystem. SIWE auth, inbox, send, setup, audit logging. Security-first design based on BaseMail architecture.
元数据
常见问题
Nadmail 是什么?
NadMail - Email for AI Agents on Monad. Register [email protected], send emails that micro-invest in meme coins, boost with emo-buy. SIWE auth, no CAPTCHA,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1511 次。
如何安装 Nadmail?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install nadmail」即可一键安装,无需额外配置。
Nadmail 是免费的吗?
是的,Nadmail 完全免费(开源免费),可自由下载、安装和使用。
Nadmail 支持哪些平台?
Nadmail 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 Nadmail?
由 Ju Chun Ko(@daaab)开发并维护,当前版本 v2.0.0。
推荐 Skills