Vendor evidence

Nightfall claims employee and device attribution, agent-to-system access mapping, allowlisted Model Context Protocol (MCP) servers, granular tool authorization, and blocking of unapproved tools and connections.

Nightfall claims controls for Cursor, VS Code, Claude, and custom Model Context Protocol (MCP) integrations including source-code and file inspection, embedded-secret redaction, server and tool allowlisting, version monitoring, request logging, and malicious-update quarantine.

Island claims unified visibility, governance, data protection, policy, audit, human oversight, usage analytics, and return on investment (ROI) reporting across browser, desktop, extension, network, embedded AI, and governed agents.

Island AI Services materials reviewed did not provide a public customer-facing product claim for automated adversarial testing, repeatable attack suites, model or agent red teaming, or release-gate evaluation.

Island AI Services materials reviewed did not establish model artifact scanning, provenance, signing, dependency or Model Context Protocol (MCP) component analysis, tamper detection, or model-registry release controls.

Island claims one policy engine across AI browser, desktop, extensions, network, prompts, outputs, agents, and more than 500 governed Model Context Protocol (MCP) integrations with prompt-injection mitigation and human checkpoints.

Island claims governed agents with defined workflows, scoped permissions, enterprise identity and policy inheritance, human-in-the-loop controls, and complete auditability.

Island claims browser, desktop, extension, network, data loss prevention (DLP), and AI policy controls that protect proprietary source code and govern AI applications, locally running agents, outputs, and developer data movement.

LayerX claims centralized policy and operational visibility across user and agent prompts, actions, identities, applications, and data exchanges in browsers, desktop apps, integrated development environments (IDEs), extensions, and on-device agents.

LayerX Interaction Security materials reviewed did not provide a public product claim for automated adversarial testing, repeatable attack suites, model or agent red teaming, or release-gate evaluation.

LayerX Interaction Security materials reviewed did not establish model artifact scanning, provenance, signing, dependency or Model Context Protocol (MCP) component analysis, tamper detection, or model-registry release controls.

LayerX claims real-time last-mile monitoring, detection, blocking, data loss prevention (DLP), and adaptive policy across prompts, agent actions, data exchanges, AI applications, browsers, desktops, integrated development environments (IDEs), extensions, and on-device agents.

LayerX claims contextual correlation of users, agents, identities, applications, and data with granular policy based on identity, data sensitivity, and risk.

LayerX claims discovery and governance of AI desktop apps, integrated development environments (IDEs), integrated development environment (IDE) extensions, browser extensions, on-device agents, prompts, actions, file transfers, copy and paste, and sensitive-data exchanges.

Aurascape claims unified discovery, inventory, policy, compliance, human and agent governance, audit conversations, identity and intent context, approved registries, and cross-channel data lineage for employee AI and built agents.

Aurascape claims predeployment adversarial guardrail tests for prompt injection and jailbreak attempts plus code-path vulnerability checks against known CVEs.

Aurascape claims discovery of agents and Model Context Protocol (MCP) servers, a vetted custom registry, tool-poisoning detection, approved-call signing, unsigned-call blocking, and project-configuration supply-chain protection.

Aurascape claims an AI Proxy and Zero Bypass Model Context Protocol (MCP) Gateway that inspect prompts, responses, intent, tool calls, parameters, data exchanges, and results; sign approved calls; and block or sanitize policy violations before execution.

Aurascape claims context-aware tool policy based on user identity, account type, agent intent, entitlement, data sensitivity, OAuth roles and scopes, and approved-call signatures.

Aurascape claims local discovery for Claude Code and Cursor, source-code and unsafe-code protections, predeployment code-path testing, Model Context Protocol (MCP) registry and signing, tool-call enforcement, and data-lineage controls from first code through runtime.