Angel Agent Guard: pre-purchase check + prompt-injection scan
Checks a shop, payment link, paid API or token before an agent pays, plus injection scans
- Pricing
- Freemium
- Platforms
- CLI, API
- Runs
- Locally
- Version
- 0.2.0
- Listing updated
- 2 days ago
About Angel Agent Guard: pre-purchase check + prompt-injection scan
Angel Agent Guard: pre-purchase check + prompt-injection scan is a freemium MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.
Agents that browse and buy read pages written by strangers, and this package gives them two checks to run first. One scans untrusted text or URLs, such as pages, emails, RAG chunks and tool results, for invisible Unicode, tag smuggling and exfiltration images, then runs a hosted scan covering 17 languages for text aimed at the model. It returns a verdict and an action: use, read only, drop or review, along with a cleaned copy to read as data.
The second check looks at where money would go. For a shop or checkout link it examines TLS, domain age, look-alike domains, redirects to payment pages on other domains, which payment provider takes the money, card fields and plain-http forms. For an x402 endpoint it inspects the offer itself, including the network, official USDC, an amount under your cap and the same host. For an ERC-20 token it looks for honeypot and rug-pull powers. Results include a risk score from 0 to 100 and a verdict of pay, caution or do not pay.
The npm package has no runtime dependencies, also ships a CLI, a Python port and adapters for the Vercel AI SDK and Coinbase AgentKit, and can pay x402 invoices automatically if you add a wallet key and a per-payment cap (default 1 USDC). The author's documentation discloses that the hosted scanner is a paid service with a free preview and three free calls per agent.
Install
Run in a terminal
claude mcp add agent-guard --env ANGEL_AGENT_PRIVATE_KEY='<ANGEL_AGENT_PRIVATE_KEY>' -- npx -y angel-agent-guardAdd to claude_desktop_config.json
{
"mcpServers": {
"agent-guard": {
"command": "npx",
"args": [
"-y",
"angel-agent-guard"
],
"env": {
"ANGEL_AGENT_PRIVATE_KEY": "<ANGEL_AGENT_PRIVATE_KEY>"
}
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"agent-guard": {
"command": "npx",
"args": [
"-y",
"angel-agent-guard"
],
"env": {
"ANGEL_AGENT_PRIVATE_KEY": "<ANGEL_AGENT_PRIVATE_KEY>"
}
}
}
}Add to .vscode/mcp.json
{
"servers": {
"agent-guard": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"angel-agent-guard"
],
"env": {
"ANGEL_AGENT_PRIVATE_KEY": "<ANGEL_AGENT_PRIVATE_KEY>"
}
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"agent-guard": {
"command": "npx",
"args": [
"-y",
"angel-agent-guard"
],
"env": {
"ANGEL_AGENT_PRIVATE_KEY": "<ANGEL_AGENT_PRIVATE_KEY>"
}
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"agent-guard": {
"command": "npx",
"args": [
"-y",
"angel-agent-guard"
],
"env": {
"ANGEL_AGENT_PRIVATE_KEY": "<ANGEL_AGENT_PRIVATE_KEY>"
}
}
}
}Configuration
ANGEL_PAYER_ADDRESSYour wallet address: its own 3 free hosted callsANGEL_AGENT_PRIVATE_KEYOptional hot-wallet key for automatic x402 payment (needs viem)ANGEL_MAX_USDCPer-payment cap in USDC, default 1.00
What it does
- Untrusted input scan
- Strips invisible characters and hidden HTML and flags text aimed at the model.
- Pre-purchase check
- Rates a shop, checkout, API or token before money moves.
- Look-alike detection
- Spots look-alike domains and cross-domain payment redirects.
- x402 offer inspection
- Verifies network, USDC contract, amount cap and host.
- Token risk check
- Looks for honeypot and rug-pull powers on an ERC-20.
- Spending cap
- Limits automatic x402 payments per call, default 1 USDC.
Good for
- Vet a checkout link before an agent pays
- Scan a web page before it enters the model context
- Verify an unknown paid API
- Check a token before approving a swap
- Drop hostile content in a RAG pipeline
Pricing
Freemium
Free preview and 3 free hosted calls per agent; paid hosted service afterwards, prices not published.
Prices change. Check angel.oroboros-design.com before you buy.
Angel Agent Guard: pre-purchase check + prompt-injection scan FAQ
Does it cost money?
The hosted checks are a paid service. The author says there is a free preview and 3 free calls per agent wallet address; prices are not published.
What languages does the injection scan cover?
The README says the hosted scan covers 17 languages.
Where does it run?
Node 18 or newer, Bun, Deno and edge runtimes that support fetch; a Python package is also listed.
Can it pay automatically?
Only if you provide a wallet key and install viem; a per-payment cap defaults to 1 USDC.
Is Angel Agent Guard: pre-purchase check + prompt-injection scan free?
Angel Agent Guard: pre-purchase check + prompt-injection scan has a free plan, with paid plans for more usage or features.
How do I add Angel Agent Guard: pre-purchase check + prompt-injection scan to Claude Code?
Run this command in your terminal: claude mcp add agent-guard --env ANGEL_AGENT_PRIVATE_KEY='<ANGEL_AGENT_PRIVATE_KEY>' -- npx -y angel-agent-guard
What are the best alternatives to Angel Agent Guard: pre-purchase check + prompt-injection scan?
The closest alternatives to Angel Agent Guard: pre-purchase check + prompt-injection scan are REA, Claude Code Ultimate Guide, Solana MCP by Vybe, Tapetide - Indian Stock Market MCP and Agent Roast.
Alternatives to Angel Agent Guard: pre-purchase check + prompt-injection scan

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling

Solana MCP by Vybe
MCP serverRemote MCP server exposing the Vybe Solana API to AI clients and code editors
- 44 tools
- Remote
- 2.8K GitHub stars
- Finance and payments

Tapetide - Indian Stock Market MCP
MCP serverIndian NSE and BSE stock market data for AI assistants through 48 MCP tools
- 17 tools
- Remote
- Freemium
- 150 GitHub stars
- Finance and payments
Agent Roast
MCP serverOffline review that flags agent changes which weaken tests, assertions or CI
Runs locally- Open source
- 4 Hacker News points
- Developer tools

Midplane
MCP serverPostgres gateway for AI agents with table grants, column masking and approval for writes
Runs locally- Open source
- 5 GitHub stars
- Databases