Clear answers.
Clearer boundaries.
Understand the products, the approach, and what to evaluate for your team.
What is Gensee Agent Execution Security?
Gensee Crate Enterprise is the safety and security control plane for autonomous AI. It brings endpoint control, cloud agent control, and central policy, logs, and investigation into one platform. Task authority, execution protection, artifact inspection and repair, and application controls help protect supported agent workflows from human mistakes, agent errors, and malicious inputs.
How does Gensee support Agent Detection and Response (ADR)?
Agent Detection and Response (ADR) focuses on detecting and responding to risky AI agent behavior. Gensee connects agent activity with task context, policy, and correlated evidence in a central dashboard, so security teams can investigate and respond to consequential exceptions. Execution controls add prevention at supported boundaries, while recovery is limited to supported local state.
Who is it for?
Teams with agents that already read sensitive information, run code, or change business systems. Operations and support are a natural starting point, with analytics, finance, and engineering workflows also in scope. The operational owner, platform or IT team, and security team define the boundaries together.
Do we need to replace our existing agents?
No. Gensee supports Claude Code, Codex, Claude Cowork, ChatGPT Work, Cursor, VS Code Copilot, Antigravity, Omnigent, and Databricks Agent Bricks. Deployment spans Mac, Windows, and Linux, on endpoints and in cloud environments, with a central log dashboard and control. Available protections depend on the harness, operating system, and application connectors. An inaccessible vendor-hosted agent needs a supported integration or a control point at the resource it uses.
Will normal work require constant approvals?
Preauthorized work should proceed under reusable policy. Routine screening and connected context are designed to reduce false alerts and unnecessary approvals. Human review focuses on consequential exceptions; model confidence alone does not authorize an action or discard high-risk evidence.
How do inspection and repair fit into execution security?
Inspection supplies evidence about code, packages, and tools. Gensee connects that evidence to task authority and execution policy. A safer version or repaired artifact must be independently validated, and approval must apply to the exact artifact used. A passing scan does not grant new permissions.
Can Gensee reverse every action?
No. Recovery applies to supported local state. It cannot unsend an email, reverse an arbitrary remote transaction, or recall disclosed data. Those effects need controls at the application or resource boundary before they occur.
Which capabilities can we evaluate today?
The architecture describes the unified platform direction. The console walkthrough and benchmarks show specific existing work; they do not establish coverage for every execution path. We scope the supported configuration and demonstrate the protections available for your workflow.
How does Personal differ from the enterprise platform?
Gensee Personal gives individual developers isolated workspaces for delegated coding tasks. The enterprise platform adds centrally managed authority, execution and application controls, shared evidence, and security review across supported organizational workflows.
What benchmark results does Gensee report?
The homepage reports preliminary internal AgentCanary and ADR-Bench results: defense rates, recall, false-positive rate, accuracy, and runtime overhead. These measurements cover specific evaluated workloads and configurations. Measure legitimate completion, missed issues, and unnecessary interruptions together in your evaluation.
Start with your workflow.
See how Gensee fits the work your people and agents do.