AgentGuard runs on your own machine — the client starts it, so there is no endpoint to ping. 192 installs a week from npm. Last commit 8 Nov 2026.
Let an AI agent pay on NEAR through a vault whose spending rules a contract enforces, not the agent.
Nothing serious here today
Today is the operative word: we check AgentGuard every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add agentguard-mcp -e 'AGENTGUARD_VAULT=<AGENTGUARD_VAULT>' -- npx -y @zaxcoraider/agentguard-mcp{
"mcpServers": {
"agentguard-mcp": {
"command": "npx",
"args": [
"-y",
"@zaxcoraider/agentguard-mcp"
],
"env": {
"AGENTGUARD_VAULT": "<AGENTGUARD_VAULT>"
}
}
}
}[mcp_servers.agentguard-mcp]
command = "npx"
args = ["-y", "@zaxcoraider/agentguard-mcp"]
env = { AGENTGUARD_VAULT = "<AGENTGUARD_VAULT>" }{
"mcpServers": {
"agentguard-mcp": {
"command": "npx",
"args": [
"-y",
"@zaxcoraider/agentguard-mcp"
],
"env": {
"AGENTGUARD_VAULT": "<AGENTGUARD_VAULT>"
}
}
}
}{
"servers": {
"agentguard-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@zaxcoraider/agentguard-mcp"
],
"env": {
"AGENTGUARD_VAULT": "<AGENTGUARD_VAULT>"
}
}
}
}Put your own value in place of
<AGENTGUARD_VAULT> (Your AgentGuard vault account, e.g. my-agent.agentguard.testnet).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
NEAR_NETWORK (testnet (default) or mainnet), AGENTGUARD_AGENT_KEY (The agent's ed25519 function-call key; without it the server is read-only), AGENTGUARD_KEY_FILE (Path to a key file to read instead of AGENTGUARD_AGENT_KEY), NEAR_RPC_URL (Override the NEAR RPC endpoint), SENTINEL_URL (Where to buy Sentinel's paid check).
Alternatives to AgentGuard
same job, measured the same wayA .NET architecture checker that enforces one C# spec and renders the same rules for coding agents.
A legal body for AI agents: a real company, a verified human and spending rules on Arc (testnet).
Search the web, news, images, videos and books, and extract page content for AI agents.
Connect accounts once. Set what agents may do per account. Enforced when the action runs.
Agents are not bots. PayClaw proves it — then lets them pay. Badge + Spend on existing Visa rails.
Trustless spending accounts for AI agents on Solana with on-chain enforced limits.
Federal contract awards and spending data as agent tools.
Search the open web and Extract Raw Content for AI agents.
AgentGuard — questions
Answers built from our own checks of this server.