paygents

Accept crypto payments in AI agent conversations. Generate MetaMask/Trust Wallet deeplinks, verify transactions on-chain, check balances, issue receipts. No custody, no backend, no API keys.

3,891 stars

Best use case

paygents is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Accept crypto payments in AI agent conversations. Generate MetaMask/Trust Wallet deeplinks, verify transactions on-chain, check balances, issue receipts. No custody, no backend, no API keys.

Teams using paygents should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/paygents/SKILL.md --create-dirs "https://raw.githubusercontent.com/openclaw/skills/main/skills/amitaybohadana/paygents/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/paygents/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How paygents Compares

Feature / AgentpaygentsStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Accept crypto payments in AI agent conversations. Generate MetaMask/Trust Wallet deeplinks, verify transactions on-chain, check balances, issue receipts. No custody, no backend, no API keys.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

SKILL.md Source

# PayGents

Accept crypto payments through your AI agent. Generate wallet deeplinks, verify on-chain, issue receipts — no backend, no API keys.

## Quick Start

Send 10 USDC on Base:
```bash
scripts/evm-payment-link.sh --to 0xRECIPIENT --amount 10 --chain-id 8453
```

Verify the payment:
```bash
scripts/evm-verify-tx.sh --chain-id 8453 --from 0xSENDER --to 0xRECIPIENT --asset ERC20 --amount 10 --blocks 50
```

Check a wallet balance:
```bash
scripts/evm-balance.sh --address 0xADDRESS
```

That's it. No API keys, no backend, no custody.

## Flow

1. Agent collects payment details (recipient, amount, chain, token).
2. Agent runs the link generator script → gets a MetaMask deeplink.
3. Agent sends the link to the user.
4. User taps → MetaMask opens with pre-filled transfer → user approves.
5. User confirms "sent" → agent verifies the tx on-chain.

## Wallet Detection

Before generating a link, the agent should know which wallet the user has. Ask once, remember forever.

**Supported wallets with native deeplinks:**

| Wallet | `--wallet` flag | Deeplink format |
|--------|----------------|-----------------|
| MetaMask | `metamask` (default) | `https://link.metamask.io/send/...` |
| Trust Wallet | `trust` | `https://link.trustwallet.com/send?...` |

**Not supported (no send deeplinks):**
- Rabby — only has in-app browser, no direct send deeplink
- Coinbase Wallet — only has dapp browser deeplink, no direct send
- Phantom — requires encrypted handshake, not a simple URL

If the user's wallet isn't supported, default to MetaMask (most common) or let them know.

Store the user's wallet preference so you don't ask again. The agent should note the preference in its local session context (e.g. memory file). The user can change or clear this at any time by telling the agent.

## Inputs Required

| Field | Required | Description |
|-------|----------|-------------|
| `--to` | Yes | Recipient address (`0x...`) |
| `--amount` | Yes | Human-readable amount (e.g. `1.5`) |
| `--chain-id` | No | Chain ID (default: `8453` Base) |
| `--asset` | No | `ETH` or `ERC20` (default: `ERC20`) |
| `--token` | No | ERC20 contract address (auto-detected for USDC on known chains) |
| `--decimals` | No | Token decimals (default: `6` for USDC, `18` for ETH) |
| `--symbol` | No | Token symbol for display (default: `USDC` or `ETH`) |
| `--wallet` | No | `metamask` or `trust` (default: `metamask`) |

## Commands

### Generate Payment Link

**ERC20 (USDC) — MetaMask:**
```bash
scripts/evm-payment-link.sh \
  --to 0x1234...5678 \
  --amount 10 \
  --chain-id 8453
```

**Native ETH — Trust Wallet:**
```bash
scripts/evm-payment-link.sh \
  --to 0x1234...5678 \
  --amount 0.01 \
  --asset ETH \
  --chain-id 11155111 \
  --wallet trust
```

Output is JSON with:
- `intent` — structured payment details
- `deeplink` — MetaMask deeplink URL
- `messageTemplate` — ready-to-send message for the user

### Verify Transaction

After the user says "sent", verify on-chain:
```bash
scripts/evm-verify-tx.sh \
  --chain-id 11155111 \
  --from 0xSENDER \
  --to 0xRECIPIENT \
  --asset ETH \
  --amount 0.001 \
  --blocks 50
```

Returns the matching tx hash if found, or "not found".

## Supported Chains

| Chain | ID | Default USDC |
|-------|----|-------------|
| Ethereum | 1 | `0xA0b86991c6218b36c1d19d4a2e9eb0ce3606eb48` |
| Base | 8453 | `0x833589fCD6eDb6E08f4c7C32D4f71b54bDa02913` |
| Sepolia | 11155111 | `0x1c7d4b196cb0c7b01d743fbc6116a902379c7238` |
| Base Sepolia | 84532 | `0x036CbD53842c5426634e7929541eC2318f3dCf7e` |

## User Message Pattern

When sending the link, always include:
1. Amount + token + chain
2. Recipient (truncated)
3. "Tap to open MetaMask and approve"
4. "Reject if recipient or amount doesn't match"

## Check Wallet Balance

Query native + major ERC20 balances across all supported chains:

```bash
# All chains at once
scripts/evm-balance.sh --address 0x1234...5678

# Single chain
scripts/evm-balance.sh --address 0x1234...5678 --chain-id 8453
```

Returns JSON with native balance + USDC, USDT, WETH, WBTC, DAI per chain.

Supported chains: Ethereum, Base, Polygon, Arbitrum, Optimism, BNB Chain, Sepolia, Base Sepolia.

No API key needed — uses public RPCs directly.

## Generate Receipt

After a transaction is verified, generate a structured receipt:

```bash
scripts/evm-receipt.sh \
  --tx-hash 0xabc123... \
  --chain-id 8453 \
  --memo "order-42" \
  --merchant "Cool Store"
```

Options:
- `--format json | markdown | both` (default: `both`)
- `--out <directory>` — save receipt files to disk (JSON + markdown)
- `--memo` — order ID or note
- `--merchant` — merchant/payee name

Output includes: status, amount, token, from/to, gas fee, block, explorer link, timestamp.

The receipt can be sent to the user as a confirmation message, saved for bookkeeping, or forwarded to a merchant.

## RPC Configuration

By default, scripts use public RPC endpoints. You can override them for privacy or reliability:

**Option 1 — Environment variables** (highest priority):
```bash
export RPC_1="https://my-private-eth-node.com"
export RPC_8453="https://my-base-rpc.com"
```

**Option 2 — Config file** (copy `config.example.json` → `config.json`):
```json
{
  "rpc": {
    "1": "https://my-private-eth-node.com",
    "8453": "https://my-base-rpc.com"
  }
}
```

**Option 3 — Public fallbacks** (default, no setup needed):
Used automatically if no env var or config is set. Public providers (e.g. `eth.llamarpc.com`) will see your wallet addresses and tx hashes.

Resolution order: env var `RPC_<chainId>` → `config.json` → public fallback.

## Security & Privacy

- The wallet is the trust boundary — agent cannot force-execute.
- Verification checks the actual on-chain receipt, not user claims.
- Never store or handle private keys.
- **RPC privacy:** If using public fallback RPCs, third-party providers will see wallet addresses and tx hashes you query. Set your own RPCs via env vars or config.json for privacy.
- **Wallet preference:** The agent may store the user's preferred wallet (metamask/trust) in its memory. This is only the wallet app name, no keys or sensitive data. The user can ask the agent to clear it at any time.

Related Skills

---

3891
from openclaw/skills

name: article-factory-wechat

Content & Documentation

humanizer

3891
from openclaw/skills

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases.

Content & Documentation

find-skills

3891
from openclaw/skills

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

General Utilities

tavily-search

3891
from openclaw/skills

Use Tavily API for real-time web search and content extraction. Use when: user needs real-time web search results, research, or current information from the web. Requires Tavily API key.

Data & Research

baidu-search

3891
from openclaw/skills

Search the web using Baidu AI Search Engine (BDSE). Use for live information, documentation, or research topics.

Data & Research

agent-autonomy-kit

3891
from openclaw/skills

Stop waiting for prompts. Keep working.

Workflow & Productivity

Meeting Prep

3891
from openclaw/skills

Never walk into a meeting unprepared again. Your agent researches all attendees before calendar events—pulling LinkedIn profiles, recent company news, mutual connections, and conversation starters. Generates a briefing doc with talking points, icebreakers, and context so you show up informed and confident. Triggered automatically before meetings or on-demand. Configure research depth, advance timing, and output format. Walking into meetings blind is amateur hour—missed connections, generic small talk, zero leverage. Use when setting up meeting intelligence, researching specific attendees, generating pre-meeting briefs, or automating your prep workflow.

Workflow & Productivity

self-improvement

3891
from openclaw/skills

Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge is outdated or incorrect, (6) A better approach is discovered for a recurring task. Also review learnings before major tasks.

Agent Intelligence & Learning

botlearn-healthcheck

3891
from openclaw/skills

botlearn-healthcheck — BotLearn autonomous health inspector for OpenClaw instances across 5 domains (hardware, config, security, skills, autonomy); triggers on system check, health report, diagnostics, or scheduled heartbeat inspection.

DevOps & Infrastructure

linkedin-cli

3891
from openclaw/skills

A bird-like LinkedIn CLI for searching profiles, checking messages, and summarizing your feed using session cookies.

Content & Documentation

notebooklm

3891
from openclaw/skills

Google NotebookLM 非官方 Python API 的 OpenClaw Skill。支持内容生成(播客、视频、幻灯片、测验、思维导图等)、文档管理和研究自动化。当用户需要使用 NotebookLM 生成音频概述、视频、学习材料或管理知识库时触发。

Data & Research

小红书长图文发布 Skill

3891
from openclaw/skills

## 概述

Content & Documentation