tddp-control-plane-audit
Analyze embedded UDP/TCP control planes such as TDDP by reconstructing framing, crypto, integrity, dispatch, authorization, and dangerous sink reachability with safe evidence-first methods.
Files2
View on GitHubSKILL.mdfirmware-reverse-eng-skill/tddp-control-plane-audit/SKILL.md
GitHubTDDP/control-plane audit
Use this skill for proprietary embedded control or debug protocols. Keep the work vendor-neutral and separate protocol acceptance from authentication.
Workflow
- Preserve hashes, extraction provenance, firmware identity, and captures. Work on derived copies; never execute firmware binaries.
- Locate service initialization, socket bind, receive loop, interface scope, feature gates, and event registration. Confirm runtime exposure separately from static binding.
- Recover framing from parser code and captures: version, type, lengths, sequence, command/category, integrity fields, encryption boundaries, padding, and bounds checks. State offsets and endianness explicitly.
- Reconstruct decode/integrity data flow. Identify algorithm, key origin, IV/nonce, covered bytes, and whether the check is keyed. Label obfuscation, integrity, and authentication separately.
- Recover command tables, indirect calls, registration conditions, and handler arguments. Classify commands as read-only, disruptive, persistent-write, or unknown.
- Trace every state-changing handler to sinks (erase, reset, reboot, config, credential, upgrade) and identify the first proven authorization gate. Treat source-IP locks and shared keys as non-authentication unless private proof is demonstrated.
- If runtime validation is necessary, choose one benign read-only command, one datagram, no retries, and capture exact bytes/timing. Never construct a destructive packet merely to prove reachability.
Reporting
Produce service/packet/table/call-graph inventories, a symbolic acceptance rule, a sink authorization table, and confidence labels: PROVEN FROM CODE, VERIFIED IN FIRMWARE, OBSERVED ON DEVICE, HIGH-CONFIDENCE INFERRED, and UNKNOWN. Do not publish ready-to-run destructive transmitters in the analysis report.