What is traffic-guard?

A high-throughput traffic and attack defense gate for incoming web traffic, shipped as zero-dependency Node.js and Python packages, with an optional TypeSafe System One tier for semantic evaluation of obfuscated attacks.

By default it runs entirely in-tree with zero dependencies in under 100 µs; when TYPESAFE_API_KEY is provided the gate elevates to jev-latest and evaluates the request across five parallel typed questions, at about 250 ms mean latency.

npm install traffic-guard for Node or pip install traffic-guard for Python; both work out of the box with no external dependencies. The Node quick start awaits trafficguard(req) and checks decision.shouldBlock.

What defense mechanisms does it implement?

Header order sequence analysis to catch forged user agents, stateless HMAC-SHA256 velocity tokens in a __trafficguard cookie, an inline 1.2 KB HashCash proof-of-work puzzle, tarpit latency and canary honeypot URLs.

Do I need to pay for keys or tokens?

No key is needed for the in-tree engine. The TypeSafe tier is optional and only used when you export TYPESAFE_API_KEY or pass an apiKey to the TrafficGuard constructor, with model jev-latest as the default.

How accurate is traffic-guard?

The README's 25-scenario benchmark reports 100% category classification, action accuracy, attack block rate and good-bot passthrough with a 0% human false positive rate, for both in-tree engines and the cloud tier.

What are the limits of the published numbers?

They are self-reported on the repo's own canonical dataset in bench/dataset.json with no independent verification, the methods are the project's own, and although the README ends with an MIT licence, the GitHub metadata in this digest lists no license.

Jev directory

CommunityJev Repos & SDKs

Traffic Guard

Zero-dependency reverse-proxy traffic classifier and bot mitigator shipped as Node.js and Python packages. The in-tree engine decides in tens of microseconds, and setting TYPESAFE_API_KEY elevates the gate to jev-latest for five parallel typed questions.

See highlights Open github.com

Stars: 1
Language: Python
Reading time: 2 min
Published by: Community
Author: hemanth
Last verified: Oct 2026

hemanth/traffic-guard

User repository on GitHub

View on GitHub

Empirical, zero-dependency reverse-proxy traffic classifier and bot mitigator in <100μs

Forks: 0
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.

Tagscommunity catalog-import logicrw security bot-mitigation

Works forJev for Classification Jev for Decision gating

Used inJev for Security & Privacy

Languages: JavaScript Python
Capabilities: Open sourceRuns locally

Highlights

In-tree and zero-dependency at under 100 microseconds: the README benchmark reports 30 µs mean latency in JS and 20 µs in Python.
Five defenses: header order sequence analysis, stateless HMAC velocity tokens, a 1.2 KB HashCash proof-of-work challenge, tarpitting and canary honeypots.
The optional TypeSafe tier sends five parallel typed questions to jev-latest when TYPESAFE_API_KEY is set, at roughly 250 ms mean latency.
Both packages expose the same shape: trafficguard(req) returns a decision with shouldBlock and reasons, or should_block in Python.
Benchmarks are runnable: node bench/bench.mjs and python3 bench/bench.py against bench/dataset.json.

Quickstart

import trafficguard from 'traffic-guard';

const decision = await trafficguard(req);

if (decision.shouldBlock) {
  return res.status(403).json({ error: 'Blocked', reasons: decision.reasons });
}

Watch out

Grounded on the truncated README and GitHub metadata reviewed 2026-10-15; the GitHub license field is empty although the README ends with MIT. All figures come from the project's own 25-case dataset in bench/dataset.json, so treat them as self-reported rather than independent.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

YouTube: Kubernetes admission control with a decision model, fail-closed.

Hacker News: SafeDep Gryph plus Jev checks for compromised coding-agent behavior.

2 pointsDetecting Compromised AI Coding Agents with Jev and GryphRead the thread on Hacker News

X: Jev and Laya on security queries

X: Testing Jev for prompt injection

From the Jev guides

Original write-ups on what this entry is for.

The Open Reimplementations: Local Models That Speak System One

Sep 22, 2026

Closed weights have not stopped the ecosystem from rebuilding the shape in the open. API-compatible servers, retrained heads on small models, and inference-time tricks that skip generation entirely.

Jev Repos & SDKs#open-models#research#local

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai