Skip to content

Evaluation Modules (A–F)

UHBS v4.2.2 defines six evaluation modules. Module D is the Safety Gate; a containment score below 95 applies an exponential penalty to the composite UHQS.


Module A: Protocol & Syntax Fidelity

Objective: Verify that the honeypot accurately implements the targeted protocol and does not reveal its emulated nature through network artifacts or timing side-channels.

Step Name Procedure
A1 Protocol FSM Inspection Issue valid, out-of-order, and invalid protocol verbs sequentially; verify exact protocol-mandated status codes.
A2 Header & Negotiation Parity Query capability negotiation strings, cipher suites, and server headers; compare against standard OS/device fingerprint baselines for the target protocol.
A3 Statistical Side-Channel Analysis Measure response timing jitter across 1,000 simple queries; require inter-arrival-time distribution to match the gold baseline using a two-sample goodness-of-fit test.

Module B: Behavioral & Stateful Realism

Objective: Confirm that the honeypot maintains consistent session state, persists dynamic modifications, and handles non-standard attacker inputs.

Step Name Procedure
B1 Cross-Session State Persistence Modify session state (file create, coil change, DB insert) and re-query; assert 100% of dynamic state changes persist.
B2 Payload Handling & Execution Send class-tailored workloads; process realistically or fail gracefully without stack traces or dropped connections.
B3 Non-UTF8 & Input Stress Fuzzing Flood with raw binary, null bytes, and improper encodings; confirm zero unhandled exceptions or unexpected resets.

Module C: Telemetry Quality & Pipeline Resilience

Objective: Ensure the decoy produces clean, structured threat intelligence (CTI) that cannot be corrupted or blinded by an attacker.

Step Name Procedure
C1 Schema Conformance Validate 100% compliance against STIX 2.1, OpenTelemetry, or ECS with accurate MITRE ATT&CK mapping.
C2 Log Injection & Parser Fuzzing Inject ANSI escapes, unescaped quotes, nested JSON, newline injections; verify sanitization and SIEM parser stability.

Force multiplier

Telemetry integrity multiplies the value of the entire security stack. Corrupted or injectable logs can blind SIEM/SOAR pipelines — Module C is a critical quality gate independent of deception realism.


Module D: Safety, Containment & Boundary Controls

Objective: Guarantee that the honeypot cannot be exploited to breach underlying infrastructure, escalate privileges, or conduct outbound attacks.

Step Name Procedure
D1 Out-of-Band (OOB) Egress Sweeps Attempt ICMP, DNS tunneling, HTTP/S, and raw TCP/UDP egress; confirm 0% unauthorized traffic exits the sandbox.
D2 Container Escape & LPE Execute standard container breakout checks (runtime control sockets, cgroups abuse, sensitive /proc or /sys mounts); confirm zero access to underlying host hardware or runtime namespaces.
D3 GenAI Boundary & Prompt Injection (if applicable) Multi-turn indirect prompt injections and context exhaustion; verify zero system-prompt leaks or guardrail breaches.

Safety Gate

A Module D score below 95 triggers the exponential multiplier \(\delta_C = (C/100)^2\), degrading the entire UHQS regardless of other module performance.


Module E: Scalability, Latency & Stress Performance

Objective: Evaluate performance under sustained, automated attack conditions.

Step Name Procedure
E1 Connection Saturation Stress Ramp workers to peak capacity; measure P50/P95/P99; ensure P95 remains below the TPS threshold (standard < 150 ms).
E2 Resource Exhaustion & Circuit Breaking Flood memory-intensive operations; verify clean circuit-breaker engagement with zero crashes or thread-blocking loops.
Metric Requirement
P95 latency < 150 ms under peak concurrent load (unless TPS overrides)
P50 / P99 Recorded for full latency distribution
Crashes 0 allowed under resource exhaustion

Module F: White-Box Static Code Audit

Objective: Analyze honeypot source and deployment artifacts for vulnerabilities discoverable prior to deployment.

Step Name Procedure
F1 Static Security & SAST Execute industry-standard SAST and container image scanners across the repository and deployment artifacts; target zero High or Critical vulnerabilities, RCE flaws, or insecure system command wrappers.
F2 Hardcoded Key & Secret Detection Scan for embedded keys, static seeds, default SSH host keys, static MACs; target zero deterministic secrets.
F3 Coverage & Unsupported Command Review Compare handlers vs. target standards; measure fully handled requests vs. unhandled fallback stubs.