MITRE ATT&CK Detection Coverage Worksheet: Map Safe Tests to Analytics and Evidence Gaps
MITRE ATT&CKDetection CoveragePurple TeamSecurity ValidationSOCDetection Engineering

MITRE ATT&CK Detection Coverage Worksheet: Map Safe Tests to Analytics and Evidence Gaps

PPayloads.live Editorial Team
2026-11-17
7 min read

Use this MITRE ATT&CK coverage worksheet to connect safe tests with telemetry, analytics, triage context, owners, confidence, and evidence gaps.

A useful MITRE ATT&CK coverage assessment measures more than how many rules are mapped to techniques. This worksheet helps security teams connect safe emulation tests with available telemetry, analytics, expected evidence, triage context, ownership, confidence, and known blind spots. Use it during a purple team lab, a tool evaluation, or a recurring SOC validation review to identify which parts of the detection chain work in practice and which still need evidence.

Overview

ATT&CK mappings are helpful for organizing detection content, but a mapped rule does not automatically prove that a technique is detectable. A rule may depend on an event source that is not enabled, a field that is lost during ingestion, or a data pipeline that delays the alert beyond its operational value. An alert may also lack the process, identity, host, or network context needed for an analyst to make a confident decision.

A practical detection coverage worksheet treats coverage as a chain of evidence:

  1. Prevention: Did a control block, restrict, or contain the test activity?
  2. Telemetry: Did the required events or records exist and reach the intended platform?
  3. Analytics: Did a rule, query, model, or correlation identify the activity?
  4. Triage context: Did the alert contain enough detail for an analyst to investigate?
  5. Response validation: Could the team perform and verify the expected response?

Record these dimensions separately. They describe control performance and evidence quality; they are not a risk score. A technique can have strong prevention but weak telemetry, or good alerting but poor response validation. Keeping those outcomes distinct makes remediation more precise.

Use only authorized, reversible tests in an isolated or explicitly approved environment. Safe payloads and emulation procedures should demonstrate observable behavior without collecting credentials, modifying production data, establishing persistence, or enabling uncontrolled lateral movement. For example, a credential access detection test can validate LSASS- or token-related alert logic without performing real credential dumping. The credential access detection lab provides a useful model for this approach.

What to track

Build one worksheet row for each technique or sub-technique in scope. Keep the scope specific enough that the test, data source, and expected evidence are unambiguous. A reusable row can include the following fields:

  • Technique scope: ATT&CK technique or sub-technique name, identifier, platform, and the behavior being tested.
  • Safe test procedure: Test name, lab or host, prerequisites, operator, approval reference, start and end time, and cleanup steps.
  • Prevention result: Allowed, blocked, contained, or not evaluated, with the control and evidence location.
  • Required data sources: Endpoint, identity, network, cloud, email, container, or application records needed to observe the behavior.
  • Telemetry result: Present, partial, absent, delayed, or unverified. Record the exact log source and representative event identifiers where appropriate.
  • Analytics: Rule name, query version, Sigma source, YARA content, SIEM correlation, EDR analytic, or hunting procedure used for validation.
  • Expected alert evidence: Fields an analyst should see, such as user, host, parent process, command line, destination, cloud principal, file hash, or message metadata.
  • Observed result: Alerted, hunted only, suppressed, misclassified, duplicated, or not detected.
  • Triage and response: Whether the alert was actionable, who owned it, what enrichment was available, and whether the response playbook was tested.
  • Confidence and blind spots: Confidence level, assumptions, excluded platforms, missing fields, known noise, and conditions that could change the result.
  • Next action: Data engineering, rule change, tuning, documentation, control change, retest, or accepted limitation.

Organize the worksheet by telemetry domain rather than relying only on an ATT&CK matrix view. The same technique may require different evidence across environments:

Coverage areaUseful evidence to trackCommon gap to test
EndpointProcess creation, command line, parent-child relationships, file and registry changes, script logging, prevention actionsEvents exist locally but are filtered, delayed, or missing from the SIEM
IdentitySign-in activity, authentication method, privilege changes, token or session details, directory changesSuccessful and failed activity cannot be correlated to the same user, device, or session
NetworkDNS, proxy, flow, firewall, connection, and protocol metadataEncrypted or east-west traffic is visible only as incomplete metadata
CloudAudit events, role changes, API calls, access-key use, resource changes, and principal contextImportant services or regions are outside the logging policy
EmailSender authentication, URLs, attachments, delivery actions, user reports, and mailbox eventsMessage telemetry is separated from endpoint and identity investigation data

Use the Sysmon event reference and the Windows audit policy baseline when endpoint evidence is in scope. For Linux systems, compare the worksheet’s required fields with the Linux Auditd detection lab.

Cadence and checkpoints

A coverage worksheet is most valuable when it records change over time. Establish a baseline after the initial test, then review it on a monthly or quarterly cadence appropriate to the environment. A quarterly review is a practical default for stable systems; higher-change environments may need monthly checks for critical techniques and data sources.

At each review, use the same test conditions where possible. Compare the test version, host or cloud scope, enabled controls, telemetry configuration, analytics version, and result timestamp. Without these details, an apparent improvement may simply reflect a different test or a changed environment.

Use these checkpoints:

  • Before testing: Confirm authorization, scope, rollback steps, expected telemetry, analytic owners, and the evidence locations to inspect.
  • During testing: Capture the exact execution window, affected asset, test identifier, prevention result, raw events, alert timestamps, and analyst actions.
  • After testing: Verify cleanup, reconcile raw telemetry with SIEM or XDR records, document alert latency, and record whether the response procedure was completed.
  • At review time: Compare open gaps, changed analytics, data-source health, false-positive observations, and retest status with the previous entry.

For rule changes, pair the worksheet with a regression process. The Sigma-to-SIEM validation workflow can help preserve known-positive and known-negative test cases while analytics evolve.

How to interpret changes

Do not summarize coverage with one percentage unless the calculation is carefully qualified. A technique marked “covered” can mean that a rule exists, that telemetry is available, or that a complete test produced a timely, actionable alert. Those are materially different outcomes.

Instead, report a small status set for each dimension:

  • Validated: The authorized test produced the expected evidence and the team completed the planned analyst or response action.
  • Partially validated: Some evidence or control behavior worked, but a field, platform, timing, or response step was incomplete.
  • Telemetry gap: The required data was not available, was not retained, or could not be reliably associated with the test.
  • Analytic gap: Relevant evidence existed, but the intended rule or hunting procedure did not identify it.
  • Context gap: An alert occurred, but analysts lacked the information needed to investigate efficiently.
  • Not tested: The technique remains outside the validated scope; do not treat it as detected or undetected.

Look for patterns across rows. If several endpoint techniques have raw events but no alerts, prioritize analytics and field normalization. If alerts fire but triage takes excessive manual enrichment, improve context and correlation. If a rule’s alert count falls after tuning, check both false-positive reduction and missed-test behavior before calling the change an improvement.

Separate platform coverage as well. A detection validated on Windows should not automatically be marked validated for Linux, macOS, cloud workloads, containers, or identity systems. For container-focused exercises, compare application, Kubernetes audit, runtime, and cloud evidence rather than collapsing them into one endpoint category. The Kubernetes runtime lab and cloud IAM lab illustrate why platform-specific evidence matters.

When to revisit

Revisit the worksheet at least quarterly and whenever a recurring data point changes. Trigger an earlier review after an EDR, SIEM, XDR, identity, cloud logging, email, or network monitoring change; a major operating system or cloud service change; a new analytic deployment; a significant ingestion or retention change; or a purple team campaign that exposes an untracked behavior.

Refresh ATT&CK technique and sub-technique identifiers, platform mappings, data-source definitions, vendor terminology, and test references after relevant ATT&CK releases and during the quarterly coverage review. Preserve the prior worksheet version so the team can distinguish a real control change from a taxonomy or naming change.

To close each review cycle, choose a small number of evidence gaps with named owners and due dates. Retest the highest-priority gaps using the same safe procedure, attach raw evidence and alert records, and update the status only after the result is reproducible. For lateral movement scenarios, use a controlled lateral movement emulation lab; for email scenarios, use the phishing detection lab. The objective is not to make a matrix look complete. It is to maintain a defensible record of what was tested, what evidence was available, what analysts could do with it, and what still needs hardening.

Related Topics

#MITRE ATT&CK#Detection Coverage#Purple Team#Security Validation#SOC#Detection Engineering
P

Payloads.live Editorial Team

Security Editorial Team

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.