ColomboAIDOCS
API ReferenceCookbook
Docs/MC-1 Security/MC-1 Security
PublishedGuidanceLAST REVIEWED · 18 NOV 2026

MC-1 Security

The consolidated platform for agent identity, policy, assessments, controlled response and evidence.

Runtime scope

MC-1 Security

One security platform connecting agent identity, policy, assessments, controlled response and evidence.

Defense Command is the operations workspace inside MC-1 Security. Identity and Guard provide authority and policy foundations; Trust retains evidence; Continuum governs the configuration-correction mission recipe. Their separate availability and permissions still apply.

Capability and availability map

  • Identity and policy — Connected foundation. Build on Agent Identity and Guard for scoped authority, policy decisions and human approvals.
  • Defense Command — Engineering preview. Bring scoped source reports, incidents, separately approved access restrictions and verification evidence into one workspace.
  • Managed assessments — Engineering preview. Review explicitly authorized uploaded MCP configuration through isolated passive inspection. Findings remain potential.
  • Governed configuration correction — Engineering preview. Bind one exact file correction to an owned target, separate review and a distinct read-only verifier. Runtime remediation is not established.
  • Evidence and assurance — Connected foundation. Use encrypted Trust evidence and scoped observations. An applied action or retained receipt is not current certification.
  • Recovery and broader protection — In development. Approved reconciliation and rollback, backup cleanup, wider integrations and production operations still require completion and qualification.

The integrated access workflow

  • Create an expiring source credential restricted to one registered agent, tenant and project.
  • Ingest normalized source reports; retain original occurrence time and durable encrypted evidence.
  • Investigate versioned detections before proposing a response.
  • Require another current human administrator to approve the frozen proposal.
  • Dispatch a leased, reversible MC-1 access quarantine under a dedicated credential.
  • Use a different verifier and a signed delegated agent proof to test the actual authorization gate.
  • Distinguish source reports, applied controls, historical observations and current assurance.

What the response controls

The supported action quarantines a registered agent at the existing MC-1 authorization gate. It does not terminate an operating-system process, isolate an endpoint, change a firewall or block an external network. A successful API acknowledgement is not containment evidence.

A current verification observation means the independently credentialed request was denied at that precise gate. Its maximum freshness window is30seconds and disappears when the owned control or proof receipt is absent.

Engineering guides

End of guide
Was this page helpful?
MC-1 Security | MC-1 Docs