KILL CHAIN CONTROL PLANEoffensive ai agent harness · wah-cf
🔒 Cloudflare Access · Zero Trust authenticated as william.o.ferguson@ethicallyhacking.space
Natural language is matched to a registry tool via run_tool. Selecting a phase below filters the menu. All runs are sandboxed, egress-gated, and audited.

Dispatch

Select a tool to configure a run.

Run / Audit log

PROTOTYPE · presentation layer only — no live container execution. Lockheed Martin Cyber Kill Chain® used as the menu taxonomy.
Authorized red-team / security-research use only. Every tool runs in an isolated Cloudflare Sandbox with a per-tool egress allowlist and target authorization gate.

Manual approval required

This tool is gated. Confirm the authorized target before dispatch — the run is recorded in the audit log either way.