C2 Matrix is a red-team/adversary-emulation reference and selection aid, not a production integration catalog or an authority for current deployment procedures. Its evaluation lab is useful as a bounded-lab design example: distinct WAN, test, victim, and sensor segments; explicit instrumentation; and repeatable test conditions. Its detection basics, beacon analysis, and JA3/JA3S discussion are useful prompts for defender telemetry coverage. The site's own about page describes a work-in-progress collection; linked catalog material and individual technical pages may age independently.
Treat these pages as design context only. Do not copy old installation steps, default credentials, or tool-specific operating instructions into a production runbook. Verify any vendor or product integration against its current official documentation and an explicitly approved test plan.
This control plane may represent authorization, lab inventory, isolation evidence, telemetry source health, and approval/audit records for a benign test on disposable, owned assets. It is not a command-and-control server and must not manage payloads, implants, listeners, beacons, redirectors, persistence, or evasion. Do not route a lab into production, OT, safety, or other critical networks. Use benign replay/simulation; any separately authorized emulation tooling remains outside this application and under the security team's control.
The Purple-team lab isolation & detection evidence control is a planning/evidence control, not an executable test. Require written scope and expiry, network-boundary review, VM inventory/snapshot, sensor and clock health, benign test manifest, correlated source events, cleanup/restore record, and a named exception owner. Beacon cadence or TLS fingerprints are contextual signals only—not attribution or proof.
Any future SIEM, EDR, NDR, hypervisor, or endpoint connector must be independently reviewed and start read-only. Before enabling it, record:
- accountable source owner, tenant, adapter/version, official API documentation, tested date, and explicit read scopes;
- approved endpoint allowlist and secret-manager reference (never credentials in UI, logs, fixtures, or demo data);
- versioned normalized event schema, immutable source event ID, source timestamp, ingest timestamp, clock uncertainty, asset identity, provenance, and data-quality state;
- bounded request/page sizes, timeouts, rate limits, retry/backoff, pagination, retention, and idempotent replay behavior;
- audit events for configuration and access, least-privilege identity, TLS verification, and a documented disable/revoke path;
- fixture-based contract tests, authorization-negative tests, malformed/stale data tests, and an explicit
configured→probed→validatedlifecycle. Never label a connector live merely because it is configured.
The UI should expose source freshness, coverage denominator, gaps, last successful read, adapter version, and degraded/error state. Do not infer absence of activity from missing telemetry. Correlate endpoint and network events only with documented time windows and uncertainty; preserve original evidence and show the derivation. Agent plans or MCP tool availability do not bypass signed policy, scoped identity, human approvals, or audit gates.
Before presenting this capability as production-ready, verify tenant isolation, source authorization, endpoint allowlisting, bounded resource consumption, stale/partial source behavior, auditability, retention/deletion obligations, backup/restore, and incident response. Live connector validation requires an owner-approved endpoint and identity; none is supplied or probed by this package. The current workload/control are planning and evidence surfaces only, not a claim that a connector or lab has been deployed.