root@thehacksparrow:~/writeups$ SYSTEM ONLINE
root@sparrow:~/writeups$ cat introduction-to-edr.md
// Blue Team & SOC

Introduction to EDR

17 Jul 2024 · 9 min read
Overview — An introductory Blue Team & SOC room on EDR (Endpoint Detection and Response). It explains why EDR reaches far beyond traditional antivirus: continuous endpoint telemetry, behavioural and anomaly detection, IOC correlation, MITRE ATT&CK mapping, and powerful response actions (host isolation, process termination, file quarantine, remote shell, and artifact collection). It closes with a simulated triage in which you investigate real detections across several endpoints.
PlatformTryHackMe
CategoryBlue Team & SOC
DifficultyEasy
RoomIntroduction to EDR

Task 1 — Introduction

An EDR continuously monitors endpoints, detects advanced threats, and provides rich context through process trees, timelines, and historical visibility. That ability to reconstruct the full story behind a detection is what sets it apart from a simple point-in-time detector.

Task 2 — Foundations and feature pillars

EDR rests on three pillars:

PillarWhat it provides
VisibilityProcess, registry, file, and user activity, with process trees to reconstruct the attack story.
DetectionBlends signatures, behaviour, and machine learning.
ResponseIsolate, kill processes, quarantine, and remote access.

Q: Which feature of EDR provides a complete context for all the detections?
A: Visibility

Q: Which process spawned sc.exe?
A: cmd.exe

EDR is host-centric: it complements network-level controls but does not replace them.

Task 3 — Beyond the Antivirus

The room's analogy makes it clear: AV is like an immigration check comparing passports against a known list (signatures), whereas EDR is like in-airport security, watching ongoing behaviour through cameras and sensors. AV often misses obfuscated PowerShell, memory injection, and unusual parent-child chains; EDR correlates behaviours, flags anomalies, and shows the full attack chain.

Q: In the analogy, what presents an AV?
A: Immigration check

Q: Which legitimate process was hijacked?
A: svchost.exe

Q: Which solution might mark this activity as clean?
A: Antivirus (AV)

Task 4 — How an EDR works

The agents (sensors) installed on each endpoint collect rich telemetry and can perform local detections. The central console correlates, enriches with threat intel, applies ML/analytics, and surfaces alerts with their severity. EDR also integrates with the wider security stack and feeds the SIEM for centralised investigations.

Q: Which component collects telemetry?
A: agent

Q: An EDR agent is also known as?
A: Sensor

Task 5 — EDR Telemetry

Telemetry is the endpoint's black box. Among the types it collects: process executions and terminations, network connections, command-line activity, file and folder modifications, and registry changes. With this data the analyst reconstructs timelines, identifies root cause, and tells benign from malicious behaviour even when each individual action looks normal.

Q: Which telemetry helps detect C2?
A: Network connections

Q: Where are Windows configuration settings stored?
A: Registry

Task 6 — Detection and Response Capabilities

Detection techniques:

  • Behavioral detection — flags unusual parent-child chains (e.g. winword.exe spawning powershell.exe).
  • Anomaly detection — flags deviations from the endpoint baseline.
  • IOC matching — correlates with known malicious hashes, domains, and URLs.
  • MITRE ATT&CK mapping — contextualises tactics and techniques.
  • Machine learning — detects complex, fileless, and multi-stage patterns.

Response capabilities: isolate the host, terminate processes, quarantine files, remote shell, and artifact collection (memory, logs, folders, registry hives).

Q: Which feature helps identify threats based on known malicious behaviors?
A: IOC matching

Task 7 — Investigate an alert on EDR (simulated triage)

The scenario asks you to triage medium/high-severity detections across several hosts using process trees, network telemetry, and threat-intel labelling. These are the findings per endpoint:

HostQuestionFinding
DESKTOP-HR01Tool launched by CMD.exe to download the payloadcURL.exe
DESKTOP-HR01Absolute path of the downloaded malwareC:\Users\Public\install.exe
WIN-ENG-LAPTOP03Absolute path of the suspicious syncsvc.exeC:\Users\haris.khan\AppData\Local\Temp\syncsvc.exe
WIN-ENG-LAPTOP03URL of the exfiltration attempthttps://files-wetransfer.com/upload/session/ab12cd34ef56/dump_2025.dmp
DESKTOP-DEV01Threat Intel label for UpdateAgent.exeKnown internal IT utility tool
Note the last case: UpdateAgent.exe is labelled as a legitimate internal tool. Threat intel doesn't only confirm what's malicious — it also helps rule out false positives and avoid wasting time on known benign activity.

Task 8 — Conclusion and analyst takeaways

The room leaves four practical outcomes:

  • Architectural understanding: agents (sensors) feed a central console with correlated alerts and contextual evidence.
  • Detection depth: EDR spots behaviours, anomalies, and complex chains that evade AV-only controls.
  • Response readiness: analysts can contain and remediate quickly using isolation, kills, quarantine, remote access, and artifact collection.
  • Investigation efficiency: process trees, command lines, network flows, and ATT&CK mapping cut time-to-triage.

Strategic advice: pair EDR with SIEM, NDR, email gateways, IAM, and DLP for defence-in-depth; keep tuned baselines and curated detection content; establish playbooks for common detections (PowerShell misuse, LOLBins, process injection, suspicious network beacons); and use retrospective hunting across historical telemetry to uncover spread, persistence, and additional compromised assets.

Beyond the room

A few concrete, independently-verifiable anchors for these concepts — public knowledge, not part of the room itself:Real EDR products follow the same agent/console model described here: Microsoft Defender for Endpoint, CrowdStrike Falcon, SentinelOne Singularity, and VMware Carbon Black are widely deployed commercial examples, alongside the open-source Wazuh and Elastic Security stacks.MITRE ATT&CK® (attack.mitre.org) is the public knowledge base most EDR vendors map their detections to. It organises adversary behaviour into Tactics (the "why" — e.g. Execution, Persistence, Defense Evasion) and Techniques/Sub-techniques (the "how"), each with a stable ID, for example T1055 Process Injection, T1003 OS Credential Dumping, and T1071 Application Layer Protocol (C2).On Windows, most EDR agents build their telemetry on top of ETW (Event Tracing for Windows) and kernel-level callbacks. Sysmon (part of Microsoft Sysinternals) is the free tool commonly used to reproduce similar process, network, and registry logging for labs and SIEM ingestion.Tools abused by attackers precisely because they are legitimate and already present on the host — like the curl.exe seen in Task 7 — are catalogued by the LOLBAS project (lolbas-project.github.io) for Windows, and by GTFOBins for Linux.The detection logic analysts write to catch these behaviours is often shared as Sigma rules, a vendor-agnostic YAML format (sigmahq.io) that can be translated into most EDR and SIEM query languages.

Key takeaways

  • EDR extends antivirus with continuous telemetry and historical endpoint visibility.
  • Its three pillars are Visibility, Detection, and Response.
  • Detection blends behaviour, anomalies, IOC matching, ATT&CK, and ML — it doesn't rely on signatures alone.
  • The agent (sensor) collects telemetry while the central console correlates it and enriches it with threat intel.
  • Triage relies on process trees, network telemetry, and threat-intel labelling to separate malicious from benign.
  • EDR is host-centric: it performs best integrated into a defence-in-depth stack alongside the SIEM.