Matrix & filters
Master the Enterprise matrix and all six filters: what each represents, how to apply them in order, and worked examples.
Matrix layout
The matrix follows MITRE Enterprise ATT&CK v20.0: 15 tactic columns from Reconnaissance through Impact. Techniques appear as cards; sub-techniques nest under parents when expanded.
Selecting a card opens the technique panel without losing your place in the matrix. Close the panel to continue browsing.
How filters combine
Six header filters narrow the same Enterprise matrix. They do not change ATT&CK v20.0 content—only which techniques are visible or highlighted.
Within one filter, multiple selections use OR logic (match any selected value). Across filters, selections combine with AND logic (technique must satisfy every active filter).
Filter dropdown options cascade: as you add filters, remaining options update so you only see values that still produce matches. The filter logic bar summarizes active filters in plain language and shows the match count.
- Search matches technique IDs (T1059.001) and names, and combines with all filters using AND logic.
- Platform, data source, and data component filters also scope telemetry rows in the Detections tab when a technique is open.
- Clear individual filters from the logic bar or reset everything with Clear.
Recommended filter order
Filters appear left to right in the header in a deliberate workflow order. You do not have to apply them sequentially, but this sequence matches how most assessments flow—from environment scope to instrumentation, adversary framing, and defense.
- 1. Platform — Environment where the technique applies, such as Windows, Linux, macOS, cloud, SaaS, containers, identity systems, or network devices.
- 2. Data source — The log source or telemetry provider referenced in detection mappings, such as WinEventLog:Sysmon, auditd:SYSCALL, AWS:CloudTrail, or macos:unifiedlog.
- 3. Data component — The telemetry or log activity useful for detecting a technique, such as process creation, command execution, authentication, file, registry, or network activity.
- 4. Tactic — The attacker's objective. It explains why the attacker performs the action.
- 5. Threat actor — A known adversary group or activity cluster reported to use specific ATT&CK v20.0 techniques.
- 6. D3FEND 1.6.0 artifact — A defensive analysis object used in MITRE D3FEND 1.6.0, such as a file, process, command, account, IP address, domain, service, or network connection.
Quick rule of thumb — Scope Platform first, validate logs with Data Source and Data Component, then narrow intent with Tactic and Threat Actor. Add D3FEND 1.6.0 Artifact when the conversation shifts from detection to preventive controls.
Platform
Environment where the technique applies, such as Windows, Linux, macOS, cloud, SaaS, containers, identity systems, or network devices.
Where adversary behavior applies in your estate: endpoint OS (Windows, Linux, macOS), cloud control planes (IaaS, SaaS, Office Suite), containers, identity systems, network devices, and more. It answers: which part of our environment could this technique target?
Shows techniques whose MITRE platform list includes at least one value you select. Select multiple platforms with OR logic—techniques matching any selected platform stay visible. Options in other filters update to stay consistent with your platform scope.
Example — Windows endpoint program — Set Platform = Windows before a detection sprint. The matrix drops cloud-only and macOS-only techniques so your team focuses on behaviors relevant to domain-joined endpoints, Sysmon, and Windows Event Logs.
Add Data Source = WinEventLog:Sysmon and Tactic = Execution for a Windows execution rule backlog. During incident response, pair Platform with Threat Actor to see adjacent TTPs on the same OS.
Data source
The log source or telemetry provider referenced in detection mappings, such as WinEventLog:Sysmon, auditd:SYSCALL, AWS:CloudTrail, or macos:unifiedlog.
Log providers MITRE names in detection telemetry mappings—not your SIEM index names, but ATT&CK labels such as WinEventLog:Sysmon, WinEventLog:Security, AWS:CloudTrail, auditd:SYSCALL, or macos:unifiedlog. It answers: which of our collected log types could detect this behavior?
Shows techniques where at least one mapped detection strategy references a selected log source. Multi-select uses OR logic. When active, the Detections tab can show only telemetry rows matching your data source selections.
Example — Cloud audit logging — Select Data Source = AWS:CloudTrail to see techniques MITRE maps to CloudTrail-style API audit events. Open T1078.004 (Cloud Accounts) and read Detections for mapped channels before writing CloudTrail-based rules.
Always pair with Platform (e.g. IaaS or AWS) so cloud techniques are not mixed with endpoint-only scope. See the Data Source Mapping guide for a full inventory workflow.
Data component
The telemetry or log activity useful for detecting a technique, such as process creation, command execution, authentication, file, registry, or network activity.
The observable activity or artifact type within a log source—process creation, command execution, file modification, network traffic, authentication, registry events, and similar telemetry categories. It answers: what kind of event should our detection logic look for?
Shows techniques whose mapped telemetry includes at least one selected data component. Narrower than data source alone: Sysmon supports many components, but not every tactic is detectable from process creation alone.
Example — Process-centric hunting — Filter Data Component = Process Creation to list techniques where defenders can detect behavior from process-start telemetry. Open T1059.001 and confirm Detections maps CommandLinePattern and ParentProcessName under the relevant analytic.
Apply after Platform and Data Source to avoid over-assuming coverage. Pair with Tactic = Execution or Persistence when grooming a rule backlog tied to EDR process events.
Tactic
The attacker's objective. It explains why the attacker performs the action.
The adversary's objective—the reason they perform an action. ATT&CK Enterprise v20.0 organizes the matrix into 15 tactic columns from Reconnaissance through Impact. It answers: which kill-chain phase are we assessing?
Shows techniques tagged with at least one selected tactic. Each matrix column is a tactic; filtering by tactic hides techniques outside those objectives. Multi-select uses OR logic across tactics.
Example — Credential Access sprint — Set Tactic = Credential Access to focus on T1003, T1110, T1558, and related sub-techniques without distraction from Exfiltration or Impact columns. Use Filtered only mode for a clean engineering backlog.
Set Platform first so the sprint matches your estate. Add Data Source or Threat Actor when scoping purple-team exercises or sector-specific threat models.
Threat actor
A known adversary group or activity cluster reported to use specific ATT&CK v20.0 techniques.
Known adversary groups or activity clusters MITRE documents as using specific techniques—nation-state affiliates, cybercrime groups, and named intrusion sets. It answers: which TTPs matter for the groups we emulate or defend against?
Shows techniques linked to at least one selected threat actor via ATT&CK relationships. Open the Threat Actor tab on any technique to see which groups MITRE reports for that behavior.
Example — Emulation scope — Select a threat actor from your latest CTI report, add Platform = Windows, and switch to Filtered only mode. Export JSON to hand facilitators a technique list aligned with the group's documented ATT&CK usage.
Combine with Tactic and Platform for purple-team tabletops. During IR, pair with search (technique ID from an alert) before expanding to adjacent techniques the same actor uses.
D3FEND 1.6.0 artifact
A defensive analysis object used in MITRE D3FEND 1.6.0, such as a file, process, command, account, IP address, domain, service, or network connection.
Defensive analysis objects in MITRE D3FEND 1.6.0—File, Process, Network Traffic, Domain Name, User Account, Service, and similar entities defenders monitor or harden. It answers: which offensive techniques connect to this type of defensive object?
Shows techniques with D3FEND 1.6.0 mappings involving the selected artifact type(s). Use when discussing preventive controls and architecture, not log-source coverage.
Example — File-centric hardening review — Filter D3FEND 1.6.0 Artifact = File to surface techniques with file-related defensive countermeasures. Open each technique's D3FEND 1.6.0 tab to discuss controls with architecture stakeholders alongside Mitigations.
Pair with Tactic during control design workshops. Use Platform to keep reviews scoped to your OS and cloud mix. Offensive scoping usually starts with Platform and Tactic; add D3FEND artifact when pivoting to defense.
Example filter combinations
Real workflows rarely use a single filter. The combinations below show how to stack filters for common security tasks. Use Filtered only mode for focused lists; use Highlight all when you need gap conversations against the full matrix.
| Scenario | Filter stack | What you get |
|---|---|---|
| Windows detection sprint | Platform = Windows → Data Source = WinEventLog:Sysmon → Tactic = Execution | Execution techniques detectable from Sysmon on Windows—open Detections for mutable parameters and channels. |
| Cloud logging gap review | Platform = IaaS → Data Source = AWS:CloudTrail → Highlight all mode | Cloud techniques mappable to CloudTrail while still seeing the full matrix for blind-spot conversations. |
| Purple-team emulation | Threat Actor = [group] → Platform = [env] → Filtered only | Exercise technique list aligned with reported group TTPs; export JSON for facilitators. |
| Active incident pivot | Search T1059.001 → Threat Actor = [suspected group] → Platform = Windows | Confirm alert context, then expand to adjacent techniques the actor may use next on Windows. |
| Architecture / hardening | Tactic = Persistence → D3FEND 1.6.0 Artifact = File → Platform = Windows | Persistence techniques with file-related D3FEND countermeasures for control design discussions. |
| SIEM coverage check | Platform = [env] → Data Source = [your sources] → Compare TIDs with Current Filter | Gap-check a TID inventory against the scoped matrix without changing filters—see the SOC & use cases framework gaps guide. |
| BAS vs detection coverage | Import → Compare Two TID Sheets (no filters required) | Compare BAS executed techniques against your documented detection coverage—matched, only-in-sheet-1, and only-in-sheet-2 on the matrix with a four-sheet Excel report. |
Filter view modes
- Filtered only: hide techniques that do not match active filters. Best for engineering sprints, purple-team scope, and exports meant to represent a defined subset.
- Highlight all: keep the full matrix visible and highlight matches. Best for coverage gap reviews, leadership briefings, and TID comparison conversations.
- Hide empty tactics: available after filtering—removes tactic columns with no visible techniques so wide matrices stay readable on smaller screens.
- The filter logic bar summarizes active filters, shows TID comparison stat cards when a spreadsheet comparison is active, and supports clearing individual filters.
Matrix display controls
Two header controls change how the matrix displays without touching your active filters.
- Expand/collapse all: one button that opens every sub-technique group across all 15 tactics at once, then folds them all back down on the next click—faster than clicking each parent technique's chevron one at a time.
- Zoom in / Zoom out: resize matrix and technique panel text through five presets—100%, 110%, 115%, 120%, 125%. Click any of the five dots to jump straight to that level, or use the buttons to step through them. Your zoom level is saved in the browser and restored on your next visit.
ATT&CK dataset version
The version dropdown in the header reloads the matrix and technique panel against a different pinned ATT&CK Enterprise release instead of the current dataset—useful for gap comparisons against an older framework version, or matching whatever release your own documentation still references.
- v20.0 (latest): the current live release, and the default on every visit.
- v19.0, v18.1, v18.0: pinned prior releases, imported the same way as the live release.
- Switching versions resets active filters and technique selection, since technique counts, sub-techniques, and mappings can differ between releases.
See the ATT&CK Version Switching interactive tutorial (Tutorial button in the matrix header) for a guided walkthrough of when to reach for an older release—gap comparisons against a prior version, or matching whatever release your own documentation still references.
Search
Search matches technique IDs (T1059.001) and names. Combine search with filters to narrow large result sets during investigations or detection backlog grooming.
During incident response, search the alert TID first, then add Threat Actor or Platform filters to expand to related techniques without losing context.