Every update to the platform -new capabilities, performance improvements, and fixes -documented as we ship.
Nebula now drives a real Chromium browser by sight, the way a human tester does. Vision-based element selection, human-fidelity input, and a live theater stream let it test modern web apps without tripping bot detection.
We open-sourced the engineering behind Nebula: the multi-agent orchestration core, attack-surface discovery, supply-chain attack simulation, and an AI-tests-AI security auditor.
Nebula now dynamically switches between conversational and operational modes based on real-time intent classification, delivering faster responses and more precise actions.
Greeting messages now skip the full classification pipeline entirely, shaving 2-4 seconds off every casual interaction and freeing compute for real tasks.
Nebula can now route tasks across multiple AI providers with automatic failover and circuit-breaker recovery - no single point of failure for inference.
Introducing the agent team: specialist security agents that coordinate via swarm, parallel, and sequential execution to cover every attack surface.
Multiple scan modes let you control depth and speed. The orchestrator now uses AI-driven phase planning to adapt strategy mid-scan.
The launch of Nebula - autonomous black-box pentesting with zero configuration. Point it at a target, and it handles the rest.
Stay Updated
Follow our progress as we build the future of autonomous pentesting. Major releases are announced on our blog and via email.
Read the blog