Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Detect DLL side-loading attacks where adversaries place malicious DLLs alongside legitimate applications to hijack execution flow for defense evasion.
.claude/skills/detecting-dll-sideloading-attacks/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | — | — |
| case-03 | ✗→✓ | ▲ Improved | — | — |
| case-20 | ✗→✓ | ▲ Improved | — | — |
| case-16 | ✗→✓ | ▲ Improved | — | — |
| case-02 | ✗→✓ | ▲ Improved | — | — |
| Concept | Description | |---------|-------------| | T1574.002 | DLL Side-Loading | | T1574.001 | DLL Search Order Hijacking | | T1574.006 | Dynamic Linker Hijacking | | T1574.008 | Path Interception by Search Order Hijacking | | DLL Search Order | Windows DLL loading priority path | | Side-Loading | Placing malicious DLL where legitimate app loads it | | Phantom DLL | DLL that legitimate apps try to load but does not exist | | DLL Proxying | Malicious DLL forwarding calls to legitimate DLL |
| Tool | Purpose | |------|---------| | Sysmon | Event ID 7 DLL load monitoring | | CrowdStrike Falcon | DLL load detection with process context | | Microsoft Defender for Endpoint | DLL load anomaly detection | | Process Monitor | Real-time DLL load tracing | | DLL Export Viewer | Verify DLL export functions | | Sigcheck | Digital signature verification | | pe-sieve | PE analysis for proxied DLLs |
Hunt ID: TH-SIDELOAD-[DATE]-[SEQ]
Technique: T1574.002
Host Application: [Legitimate signed executable]
Sideloaded DLL: [Malicious DLL name and path]
Expected DLL Path: [Where DLL should legitimately be]
DLL Signed: [Yes/No]
App Location: [Expected/Anomalous]
Host: [Hostname]
Risk Level: [Critical/High/Medium/Low]| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +26 percentage points is the difference between those two pass rates over the 23 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.