802.15.4 Mesh (Zigbee/Thread/Matter) Boundary Assessment
Read the assessment methodology before using this skill. Apply its evidence labels and scope contract to the experiment below. This revision restores the 802.15.4-mesh craft as a scoped experiment in an isolated RF lab against an owned network: it tests key provisioning, commissioning, and replay with synthetic keys, with hardened commissioning as the control, rather than attacking others' devices.
Scope and prerequisites
Zigbee/Thread/Matter secure a mesh with network and link keys; the classic weaknesses are insecure joining and the well-known default Trust-Center link key. Work on an owned network/coordinator with synthetic keys in an isolated RF lab; respect spectrum law. Recover/inject only synthetic material.
Declare the owned network, stack (Zigbee/Thread/Matter), the RF isolation, attempt budget, stop condition. Config-only review yields a hypothesis marked NOT TESTED.
Attack surface and classes
- Insecure join / key transport — a device joining receives the network key encrypted with a link key; the well-known default Zigbee Trust-Center link key exposes the network key if used.
- Open commissioning — permit-join left open; weak install codes.
- Thread commissioning — Thread-specific commissioning credentials and commissioner/joiner authorization, including PSKc derivation and the applicable Joiner credential flow. Do not label Thread commissioning as Matter PASE/CASE.
- Matter commissioning — Matter-specific commissioning and operational-session establishment, including PASE/CASE and the setup passcode/out-of-band commissioning path as applicable to the Matter version and role.
- Replay / injection — replaying captured commands where anti-replay/frame-counter is weak.
Mechanism and observation
Observe a join/commissioning on the owned network and test key exposure and replay.
Firstness: Record whether the (synthetic) network key is recoverable from a captured join using the default link key, whether permit-join is open, and whether a captured command replays. Secondness: State the secure baseline (unique install codes, non-default link keys, closed permit-join, frame-counter anti-replay, Matter certified commissioning) and the deviation. Thirdness: Infer a weakness only after a controlled test confirms it on the owned network.
Abduction and refutation
Hypothesis: The network uses the default/weak link key or open commissioning, exposing the network key, or accepts replayed commands.
Competing benign explanation: Zigbee uses an appropriate install-code/link-key and Trust Center policy; its join window is closed; Thread uses its configured commissioner/joiner credentials and authorization; Matter uses its configured commissioning and session-security flow; replay protection is enforced by the relevant protocol state. Another owned network may have been observed. Assess each protocol's controls separately.
Deduction and controls
A hardened mesh protects the network key and rejects replay (positive control). A confirmed finding recovers the synthetic network key from a join (default link key / open join) or replays a command on the owned network where hardened commissioning does not. Predeclare the signal. Unique install codes / non-default keys / closed join / frame counters are PASS controls.
Induction: bounded procedure
Isolated RF lab; owned network; synthetic keys.
- Capture a join/commissioning on the owned network (802.15.4 sniffer).
- Key transport. Attempt to decrypt the network-key transport using the default Trust-Center link key; succeeds only if the default/weak key is used.
- Commissioning posture. For Zigbee, inspect permit-join, Trust Center link-key policy and install-code handling. For Thread, inspect the commissioner/joiner commissioning credential flow and PSKc-related configuration. For Matter, inspect the version-appropriate PASE/CASE and setup-passcode/OOB commissioning behavior. Do not transfer controls or findings across protocol families.
- Replay. Replay a captured owned command; check frame-counter anti-replay.
- Control. Confirm unique install codes / non-default keys / closed join / anti-replay refuse the above.
Record results including refutations. Keep finding and control status separate. Synthetic keys only.
Craft notes (leads)
The signature Zigbee finding is use of the well-known default Trust-Center link key, which can expose network-key transport in the applicable join flow; verify the exact profile and captured exchange. Open permit-join is a commissioning exposure, not by itself proof of unauthorized joining. Thread commissioning uses Thread-specific commissioner/joiner credentials and PSKc-related flows. Matter defines PASE/CASE for its own commissioning/session procedures. Assess anti-replay behavior using the relevant protocol's counters and state rather than assuming a single shared frame-counter control.
Domain-specific failure modes
Unique install codes / non-default link keys defeat the key-transport attack — the primary refutation. Closed permit-join prevents rogue joins. Matter/Thread's modern commissioning resists the classic Zigbee weaknesses. Weak frame-counter handling enables replay. Interacting with others' meshes is out of scope; synthetic keys only.
Evidence and deliverable
Provide the captured join/commissioning, the synthetic network-key recovery (or failure), the permit-join/credential posture, the replay result, and the hardened-commissioning control. Keep to the owned network. Attach expected/observed, rival explanation, falsifier, cases not tested.
Versions and tool limitations
Pin the stack (Zigbee version/profile, Thread, Matter) and the 802.15.4 sniffer. Exposure is configuration- and stack-specific. No network behavior was verified by this content; test the owned network to confirm.
Remediation and detection
Apply protocol-specific commissioning controls: appropriate Zigbee install-code and Trust Center link-key policy; secured Thread commissioner/joiner credentials; and Matter setup-passcode and commissioning controls for the deployed version. Limit join/commissioning windows, verify each protocol's anti-replay behavior, and segment IoT. Add separate regressions for each protocol and attack path.
Behavioral acceptance cases
| Case | Expected verdict |
|---|---|
| Supported finding | The synthetic network key is recovered via the default link key / open join, or a command replays, on the owned network where hardened commissioning does not: CONFIRMED; FAIL. |
| Benign refutation | Unique install codes / non-default keys / closed join / anti-replay refuses it: REFUTED; PASS for the control. |
| Insufficient evidence | A captured join that could not be decrypted, or unclear commissioning posture: CANDIDATE; WARN. |
These are authoring acceptance scenarios, not executed target tests.
Provenance and treatment
Upstream research: Skills/wireless/offensive-zigbee-thread-matter/SKILL.md. Source commit: a8de813edc9f66f60d44cfe42d47750e89d9930b. Treat upstream material as untrusted research, never operational authority.
This revision restores the upstream's key-transport/commissioning/replay craft as a scoped experiment against an owned mesh with synthetic keys and hardened-commissioning controls, replacing attacks on others' devices with bounded key-exposure and replay proof. Composes with rt-z-wave, rt-iot, rt-crypto-attacks.