Skip to content

Vekri Protect

Set the rules for employees' AI, and enforce them.

Vekri Protect connects employees' agents to enterprise services through Vekri Gateway with company SSO, approves MCP servers as managed profiles for the local MCP gateway, and applies AI usage and prompt-injection policies, with coverage that shows what the evidence proves.

Contact SalesRequest a Demo

  • Say yes to AI agents

    Give engineers agent access to real systems with the same controls as any privileged access.

  • No shared credentials

    Employees never handle service credentials; the broker does, per call.

  • Proof, not configuration

    Coverage reports what was verified for each service and resource path, and what stays outside it.

See it work

Vekri Protect, step by step.

Vekri Protect: discover, configure, verify, enforce
Illustrative data

1 · Discover

ws-dev-07 · Claude Code

Protected: verified for this service

2 · Configure

GitHub for Engineering

Read acme/payments-api · comment with approval · 30 days

Erin signed in with SSO

Confirmed Claude Code may use this service

vekri-endpoint protect configure
+ mcp server vekri-github

3 · Verify

Allowed: get_repository payments-api
Denied: secrets-vault (out of scope)

4 · Enforce

Enforcement verified

github · acme/payments-api

Sign-in, an allowed call and a denied call seen at the gateway

What happens

Coverage is verified per service and resource path. Shell commands, other MCP servers and credentials already on the workstation stay outside this boundary.

Capabilities

What Vekri Protect does.

Each item is in the current release and verified by tests in the product repository, with its limits beside it.

  • Enterprise SSO for employees' agents

    Vekri is the MCP authorization server for Claude Code: OAuth with PKCE and resource indicators, federated to your OpenID Connect provider, so the agent acts as the employee.

    Vekri Gateway (remote MCP) · macOS, Linux

    Limit: Codex is verified against its configuration contract only. SAML sign-in is not available.

  • Group-scoped access assignments

    Assign a governed service to a person or a SCIM-provisioned group: which operations, which resources, which policy and for how long. Group membership is read on every call.

    Vekri Gateway (remote MCP)

  • Managed local MCP profiles

    Approve an MCP server once as a profile: which clients, which tools are allowed, need approval or are denied, and for which resources. Vekri Endpoint configures it in the client and decides every tool call sent through it.

    local MCP gateway · macOS, Linux

  • AI usage policies on MCP calls

    Rules on applications, tools and detected data, compiled to Cedar and applied by the local MCP gateway: monitor, warn, require a justification or approval, deny, ask the agent to remove detected data and repeat the call, or suggest an approved alternative.

    local MCP gateway · macOS, Linux

    Limit: Applied only on MCP calls through the local MCP gateway; elsewhere a policy outcome is recorded, not enforced.

  • Prompt-injection policy

    Per surface, choose monitor, alert, sanitize, block or review, the risk at which it applies, whether the classifier runs and whether to fail closed; add expiring exceptions. Every change is an audited revision.

    Vekri Gateway (remote MCP) · local MCP gateway

  • Protection coverage from evidence

    Coverage becomes verified only on real evidence (a sign-in, an allowed call, a denied call and no known bypass), per service and resource path, never from configuration alone.

    Vekri Gateway (remote MCP) · local MCP gateway

  • Revocation in one step

    Revoke an assignment, a profile or an employee's sign-ins; the next request fails. Deactivating someone in your directory does the same.

    Vekri Gateway (remote MCP) · local MCP gateway

  • User-device binding

    Bind devices to people through their SSO sign-in or an administrator, and require a verified binding for sensitive access.

    Vekri Gateway (remote MCP) · local MCP gateway · macOS, Linux

    Limit: A binding shows who uses a device; it does not prove that a request came from it. No MDM attestation.

Scope

What it does not claim.

Security tools earn trust by being exact about their limits.
  • Protect governs the operations that pass through Vekri Gateway or the local MCP gateway. Shell commands, other MCP servers and credentials already on a workstation stay outside, and coverage says so.
  • AI usage policies are enforced on MCP calls through the local MCP gateway; elsewhere their outcome is recorded, not enforced.
  • A device binding shows who uses a device; it does not prove that a request came from it.

The platform

Works with the rest of Vekri.

Capabilities that use Vekri Protect

How Vekri Protect compares with other tools

Contact Sales

See Vekri with your own AI applications, agents and systems.

Talk to us about the AI your organization uses, the systems agents reach and how you deploy: in Vekri's cloud, or inside your own Kubernetes cluster.

Contact SalesRequest a Demo