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.

Install

pip install frisk-screen

Usage

from frisk import Client, Policy

client = Client()  # lite mode, no key required

result = client.screen(
    "0x9a3f1b2c3d4e5f60718293a4b5c6d7e8f9a0bc12",
    endpoint="https://api.seller.x402/quote",
    amount=2.50,
    asset="USDC",
    policy=Policy(max_per_call=5.00),
)

if not result.allowed:
    print(result.verdict, result.trust_score, result.reasons)
    # hold the payment for review

Lite mode vs. hosted

Lite mode catches obvious problems — malformed counterparties, payTo swaps, insecure endpoints, policy violations, and a small seed blocklist. For reputation history and continuously updated threat intelligence, pass an API key:

client = Client(api_key="...")

Design

• Zero runtime dependencies. The SDK uses only the standard library.
• Advisory, not in-path. Frisk never holds your funds or blocks a payment itself; it returns a verdict and your code decides.
• Typed. Ships with type hints (py.typed).

Development

This project uses uv.

uv sync
uv run pytest

License

MIT — see LICENSE. Security disclosures: see SECURITY.md.