Frisk

Pre-transaction risk screening for autonomous AI agents.

Before your agent pays an x402 seller or calls an unfamiliar tool, ask Frisk whether the counterparty is trustworthy and whether the transaction fits your policy. Frisk returns a verdict — allow, review, or block — with a trust score and human-readable reasons. It is advisory: your agent stays in control of the decision.

import { Client } from "frisk-screen";

const client = new Client(); // lite mode, no key required

const result = await client.screen("0x9a3f1b2c3d4e5f60718293a4b5c6d7e8f9a0bc12", {
  endpoint: "https://api.seller.x402/quote",
  amount: 2.5,
  asset: "USDC",
  policy: { maxPerCall: 5.0 },
});

if (!result.allowed) {
  console.log(result.verdict, result.trustScore, result.reasons);
}

Surfaces

Both SDKs expose the same model: a Client with a screen() call, a lite mode that runs locally with zero dependencies, and an optional hosted mode for reputation history and live threat intelligence.

MCP server

For agents that cannot import a library, and for asking the question interactively, the same checks are available as an MCP server exposing one tool, screen_payment:

{
  "mcpServers": {
    "frisk": {
      "command": "npx",
      "args": ["-y", "frisk-mcp"]
    }
  }
}

No API key and no account: with no configuration it screens entirely on your machine. It is listed in the MCP registry as dev.tryfrisk/frisk.

An MCP tool runs only when a model chooses to call it, so a check the model can skip is a weaker guarantee than the same check on the code path that signs the payment. Where the money actually moves, prefer the SDK. Details in mcp/.

Lite mode vs. hosted

Lite mode catches obvious problems — malformed counterparties, payTo swaps, insecure endpoints, policy violations, and a small seed blocklist — without a network call. The hosted API (https://api.tryfrisk.dev) adds reputation history and continuously updated threat intelligence.

What a verdict covers

Frisk screens who you are paying. An allow means nothing disqualifying was found by the checks you gave it enough information to run: the counterparty parses as an address and is not on the blocklist, the payTo the endpoint asked for matches the counterparty you named, the endpoint is served over HTTPS, and the amount and asset fall inside the policy you supplied. A check whose input you leave out does not run and does not fail — omit observedPayTo and no payTo comparison happens. In hosted mode an allow also means no adverse reputation history was found.

It says nothing about what comes back. Whether the response matches the shape you expected, contains the data you paid for, or is worth the price is a separate question, and Frisk deliberately does not answer it. Verifying the response contract is worth doing; it belongs after the call, on the buyer's side, against the buyer's own definition of a satisfactory answer.

An allow is not a claim of safety in general either. In lite mode confidence is always low, because the checks are structural: a counterparty with no history and no defects screens the same as one with a long clean record. The verdict is one input to your decision, which is why it is advisory.

Design principles

• Advisory, not in-path. Frisk never holds your funds or blocks a payment itself; it returns a verdict and your code decides.

• Zero runtime dependencies. The TypeScript SDK is built on the platform fetch API (Node, Bun, Deno, Workers, browser); the Python SDK uses only the standard library.

• Typed. Both SDKs ship with full type information.

Contributing

See CONTRIBUTING.md. Security disclosures: SECURITY.md.

License

MIT

The hosted API at api.tryfrisk.dev is additionally governed by the Terms of Service.