ATLAS CREW Security
CLOSED-LOOP SECURITY MEASUREMENT

What does your defense actually do?
Now you can measure it.

Most security tooling can't tell you whether a WAF actually works. Atlas Crew Security can. Crucible attacks. Synapse decides. Chimera receives. Crucible's ground-truth assertions evaluate whether what happened matches what should have happened — given Synapse's configuration. Four open-source products, one closed-loop measurement system.

HOW IT WORKS

A closed-loop system for measuring what your defenses actually do.

Three flagship products form a measurement loop. Bridge orchestrates them. Each request carries an expected outcome; the assertion fires when actual matches expected — or doesn't. That's the whole point: without a defender in the path, half the assertions break. Synapse is the defender. Crucible is the assertion engine. Chimera is the canvas they share.

Bridge SERVICE ORCHESTRATOR — ORCHESTRATION SUBSTRATE — REQUESTS RESPONSES CRUCIBLE ADVERSARY EMULATION ATTACKER + ASSERTIONS SYNAPSE DETECT. DECIDE. FIRE. INLINE DEFENDER BLOCKED CHIMERA VULNERABLE TARGETS STANDARDIZED TARGETS
— Crucible Runs adversary scenarios with ground-truth assertions. Each request carries an expected outcome; the assertion fires when actual matches expected — or doesn't.
— Synapse Inline policy enforcement at line rate. Sub-millisecond detection, embedded decisions, fleet intelligence. The defender that makes the assertions meaningful.
— Chimera 480+ vulnerable endpoints across 25 industry verticals. Standardized targets that give Crucible's scenarios a consistent canvas to write against.

— Bring your own WAF. Crucible's assertions work against any defender. Use Synapse for the integrated reference path, or swap it for Cloudflare, AWS WAF, or your own stack — Bridge handles the routing.

THE PLATFORM

Four products. One measurement system.

Three flagship products at peer level — each useful on its own, integrated into a closed loop. Bridge orchestrates the lab. Each product page covers install, architecture, and the underlying engineering.

— Bring your own WAF.

Atlas Crew Security ships with Synapse as the integrated reference defender, but Crucible's ground-truth assertions don't care what's in the path. Point Bridge at Cloudflare, AWS WAF, F5, or your own stack and use Crucible's scenarios to measure what those defenders actually catch — against the same standardized Chimera targets, with the same pass/fail rigor.

GET STARTED

One command to a working lab.

Bridge handles the lifecycle. Crucible runs scenarios. Synapse sits in the path. Chimera waits for traffic. Local-first, no signups, no telemetry, no SaaS dependency.

~/lab — atlas crew security
$ npx @atlascrew/bridge up ↳ pulling synapse v0.4.1 ............................. ready ↳ pulling chimera v0.6.0 ............................. ready ↳ pulling crucible v0.2.0 ............................ ready ↳ wiring traffic: cruciblesynapsechimera ↳ console: http://localhost:8080 $ bridge run-scenario kafka-takeover ↳ scenario: kafka-takeover · 14 steps · MITRE T1190, T1078 ↳ executing ......................................... complete ↳ assertions: 11 passed · 3 failed · 0 errored unauth API access blocked at policy edge.api.auth JWT replay rejected (synapse fingerprint match) privilege escalation succeeded — FAIL expected: block · actual: allow remediation: enable rule edge.role.assignment $ _