Skip to main content

Academic report · Malware analysis

STX RAT Malware Analysis

~2 min read
Individual malware analysis and report author
Completed · 72-page report

An evidence-led analysis of an installer-themed STX RAT sample that distinguishes direct local evidence, public sandbox observations, reputation data, and family-level research.

Malware AnalysisThreat IntelligencePE AnalysisSandboxing
Cover of the STX RAT malware analysis academic report

Investigation workflow

Evidence before attribution

The report moves from reproducible local facts to externally observed behavior, then bounds the final assessment.

  1. 1

    Acquire and hash

    Identity and chain of custody

  2. 2

    Static analysis

    PE structure, strings, imports, entropy

  3. 3

    Sandbox evidence

    Controlled public execution records

  4. 4

    Reputation correlation

    Multi-source family context

  5. 5

    Bounded assessment

    Supported and unconfirmed claims

Swipe horizontally →

Question and evidence standard

The goal was not to repeat family-level claims, but to determine what this specific installer-themed sample could support. Every conclusion is labeled by evidence source, and behavior is described as locally observed only when local artifacts prove it.

Analysis workflow

The investigation combined safe local inspection with external execution and reputation evidence.

  • Verify identity, hash, file type, and chain of custody
  • Inspect PE metadata, imports, strings, entropy, sections, and overlay data
  • Correlate peframe indicators for packing, anti-debugging, crypto, registry, token, and file operations
  • Review ANY.RUN, Hatching Triage, CAPE, and FileScan execution evidence
  • Cross-check MalwareBazaar, VirusTotal, and Hybrid Analysis

Key finding

The strongest supported assessment is an installer-themed STXRAT delivery or loader artifact. Evidence supports packing, setup-wrapper behavior, a decoy hardware-monitor presentation, and PowerShell-staged execution.

Analytical restraint

Live C2 contact, durable persistence, credential theft, and recovery of a final payload were not confirmed for this sample. Recording those limits is part of the result, not a gap to hide.