Blog / Notation & Tooling

Dual-Layer Notation in the TLCTC Framework

Strategic simplicity for people, operational precision for machines. #X and TLCTC-XX.YY name the same ten clusters, so one classification travels from the board report to the SIEM rule and back.

Bernhard Kreinz • • updated • Loading... • TLCTC v2.5.1
Revised 2026-09-15 for TLCTC v2.5.1

Cut to the essentials. Cluster wording is quoted verbatim from the v2.5 dictionary, the operator set and the sub-cluster catalogue are linked rather than restated, and the worked classifications live in the posts that own them.

One Vocabulary, Two Layers

A threat taxonomy earns its keep when the boardroom, the SOC and the development team name the same threat the same way. TLCTC does this with two notation layers over one set of ten clusters: a strategic layer written for people and an operational layer written for machines. Same clusters, same numbers, different jobs.

Strategic: #X and attack paths

  • #1 to #10 name the ten clusters
  • The arrow shows progression: #9 → #3 → #7
  • (#X + #Y) only when the order is genuinely unknown; if one step enables the next, write it sequentially
  • →[Δt=24h] annotates attack velocity between steps
  • Bridge clusters #8, #9, #10 carry a boundary crossing: ||[context][@Source→@Target]||
  • The path ends where the consequence begins: + [DRE: C], [DRE: I], [DRE: A], with refinements

Built for incident narratives, threat modeling and executive discussion. The full operator set is core paper §7.

Operational: TLCTC-XX.YY

  • XX is the cluster number, zero-padded: 01 to 10
  • .00 is the cluster itself: TLCTC-08.00 = #8
  • .Y0 is a sub-cluster, a vector class within the cluster; strategic shorthand #X.Y
  • .YZ (ones digit not 0) is a refinement within that sub-cluster
  • Up to 81 positions per cluster, all sorting cleanly as database keys

Built for SIEM rules, detection engineering, CTI exchange and CWE/CVE enrichment. The catalogue is the operational enumeration.

The one rule that holds the layers together (R-SUBCLUSTER)

A sub-cluster must reach the same generic vulnerability as its parent cluster through a different vector. If a candidate needs a different generic vulnerability, it is a different cluster, not a sub-cluster. The ten clusters are frozen with the core paper; the enumeration beneath them is the living, community-extensible part of the framework.

Why Two Layers

  1. One number, end to end: the cluster in the board report is the cluster in the detection rule
  2. Machine-readable: the strategic notation has a formal ABNF grammar and a Layer 3 JSON schema, so paths can be parsed, validated and exchanged
  3. Extensible below a stable top: new vectors become sub-clusters and refinements; the ten clusters stay put
  4. Portable: both layers map onto MITRE ATT&CK, CWE, CVE, NIST CSF and ISO 27001 without changing the number

An Example: Physical Attack

Strategic: #8 in a path

#9 ||[human][@External→@Org]|| → #8 ||[physical][@External→@Org]|| → #7 + [DRE: C]

Social engineering talks the attacker past reception, physical access to a workstation follows, and the implant is recorded as #7 at the moment it executes. The chain ends where the Data Risk Event begins.

Operational: TLCTC-08.00

The top-level cluster, defined in the dictionary as: “Unauthorized physical interaction with or interference to hardware, media, interfaces, or signals—via direct contact or exploitation of physical phenomena/emanations.”

Generic vulnerability: “Physical accessibility of infrastructure and the exploitability of physical-layer properties.”

Wording as in tlctc-framework.v2.5.json, the single normative source for cluster definitions.

Its two sub-clusters:

Operational IDStrategicVectorWhat it covers
TLCTC-08.00#8—The cluster itself (reserved)
TLCTC-08.10#8.1MechanicalPhysical contact with matter: tampering, theft, intrusion
TLCTC-08.20#8.2SignalEnergy propagation without contact: TEMPEST, acoustic, environmental

Both vectors reach the same generic vulnerability, physical accessibility, which is what makes them sub-clusters rather than clusters. The split is physics, and it maps to two control regimes: physical access control on one side, shielding and emission control on the other. An engineer can scope work by sub-cluster (“restrict server-room access for TLCTC-08.10, add emission monitoring for TLCTC-08.20”) while the board hears #8.

The original post numbered these TLCTC-08.01 and TLCTC-08.02. Since v2.5 the ones digit is reserved for refinements, so the vector classes sit on the tens digit.

The Same Chain in Both Layers

The two layers are two renderings of one path. Two examples from other posts, in strategic and operational form; the classification reasoning is in the linked posts, not repeated here.

A CVE: Hyper-V VSP, CVE-2025-21333

… → #2 |[privilege][@local_user→@SYSTEM]| + [DRE: C, I, A]

Operational: TLCTC-02.20 (#2.2, the core-function vector of Exploiting Server). No #7, because no foreign code runs; the intra-system operator is an annotation and does not change the cluster. Why this CVE is #2 · CVE/NVD enrichment.

A campaign: Emotet to ransomware

#9 ||[human][@External→@Org]|| →[Δt=hours] #7 →[Δt=minutes] #7 →[Δt=days] #1 →[Δt=minutes] #4 →[Δt=minutes] #7 + [DRE: C, Ac]

Operational: TLCTC-09.00 → TLCTC-07.00 → TLCTC-07.00 → TLCTC-01.00 → TLCTC-04.00 → TLCTC-07.00. Credential theft and credential use are separate steps, and “ransomware” is never a step: the payload is #7, the outcome is Ac. The STIX & TLCTC integration post.

Key takeaway. #X for people, TLCTC-XX.YY for machines, the same ten numbers at both ends. A single classification travels from the incident report to the SIEM rule and back to the board without translation.

Where the Details Live