Cynative
Cynative is a foundation-model lab for cybersecurity building a sovereign cyber model that runs under the control of the team it protects. Its open-source framework lets security engineers deploy read-only agents with live access to code, cloud and runtime infrastructure while keeping model context and findings inside their environment.
At a Glance
- Security engineering teams in high-stakes environments
- Enterprises requiring data sovereignty or air-gapped deployment
- Teams operating AWS, GCP, Azure, Kubernetes, GitHub and GitLab infrastructure
- Organizations needing live code-to-cloud-to-runtime security research
- +1 more
AI Tools by Cynative
(1)
Cynative
AI Security Agent Framework CLI
Discussions
No discussions yet
Be the first to start a discussion about Cynative
Latest News
Cynative published Securing AI Agents on AWS: One HTTP Tool, One Action Gateway, detailing its generic HTTP tool and operation-level AWS IAM authorization gateway.
Cynative published Building Security Agents That Cannot Escape Their Trust Boundary, explaining the sandbox and action-gate architecture for live read-only infrastructure access.
Help Net Security profiled Cynative as an open-source deep-research security agent with read-only-by-default access to code, cloud and runtime.
Products & Services
Open-source Apache-2.0 single-binary framework/CLI for live, read-only access to code, cloud and Kubernetes, with 45 reviewed built-in agents and operator-selected frontier models.
Reviewed methodologies spanning AWS, Azure, GCP, GitHub and Kubernetes, including privilege escalation, exposure, secrets, supply chain, detection, inference, Kubernetes control-plane and hardening investigations.
A cybersecurity-specialized foundation model intended for customer-cloud or air-gapped deployment, grounded and tunable on customer data, with reasoning and evidence attached to answers.
Market Position
Cynative differentiates from hosted SaaS security products such as Palo Alto Networks, CrowdStrike and Wiz, and from AI-native challengers, by running as a self-hosted binary with live read-only infrastructure access, a sandbox and an external action gate. It positions itself against coding agents and MCP servers through sandbox research, evidence verification, credential scoping and fail-closed operation-level authorization.
Leadership
Founders
Shaked Zin
CEO and co-founder; previously led engineering at Palo Alto Networks after its acquisition of PureSec, which pioneered cloud-native runtime and IAM security.
Yuri Shapira
CTO and co-founder; previously an architect/principal software engineer and engineering leader at Palo Alto Networks after the PureSec acquisition, and previously a PureSec security researcher.
Executive Team
Shaked Zin
Co-founder and CEO
Former Palo Alto Networks engineering leader following its acquisition of PureSec.
Yuri Shapira
Co-founder and CTO
Former Palo Alto Networks architect/principal software engineer and engineering leader following the PureSec acquisition; previously a PureSec security researcher.
Founding Story
Cynative was started to build a cybersecurity model that belongs to the team it protects, runs where its infrastructure already is, and can be grounded in the team's code, cloud and findings. The founders also built an open-source agent framework for live infrastructure research with a read-only trust boundary, avoiding hosted-model data-sovereignty and synchronization tradeoffs.
Business Model
Revenue Model
The security-agent framework is free and open source under Apache-2.0. The model is in development and the website offers an early-access design-partner program; no public paid pricing or revenue figures were identified.
Target Markets
- Security engineering teams in high-stakes environments
- Enterprises requiring data sovereignty or air-gapped deployment
- Teams operating AWS, GCP, Azure, Kubernetes, GitHub and GitLab infrastructure
- Organizations needing live code-to-cloud-to-runtime security research
- Security teams evaluating sovereign or self-hosted AI agents
- Cloud privilege-escalation and attack-path analysis
- Network exposure and public storage/datastore discovery
- Credentials, secrets and IAM assessment
- CI/CD and software-supply-chain investigations
- Detection and audit-coverage analysis
- Kubernetes control-plane and workload review