hardware-security

Warn

Audited by Socket on Sep 19, 2026

7 alerts found:

Securityx3Anomalyx4
SecurityMEDIUM
SKILL.md

This skill is internally coherent and not obviously malicious, but it is a high-risk offensive security skill for AI agents. It enables real hardware attack workflows, firmware extraction, credential discovery, RFID/NFC cloning, and evasion techniques; there are no clear exfiltration or installer-redirection signs, so the concern is dangerous capability rather than deception.

Confidence: 90%Severity: 88%
SecurityMEDIUM
payloads.md

The fragment is a high-risk offensive embedded-device security testing guide. It contains numerous commands capable of extracting credentials and secrets, cloning or replaying access credentials, bypassing hardware protections, injecting faults, disrupting wireless networks, and modifying target execution. These actions may be appropriate in an authorized laboratory or assessment but are dangerous outside that context. The fragment does not itself show package-install execution, persistence, covert exfiltration, or obfuscation, so malware intent is not established; however, inclusion in an unrelated software supply-chain package would be highly suspicious and should be investigated with the surrounding file and package metadata.

Confidence: 97%Severity: 91%
SecurityMEDIUM
guides/side-channel-attack-guide.md

The fragment is an explicit side-channel attack and countermeasure-evaluation toolkit/documentation. It includes code and workflow capable of collecting traces, observing AES cache accesses, and recovering and printing cryptographic keys. It contains no evident malware behaviors such as network exfiltration, persistence, destructive actions, or obfuscated payloads. The primary risk is dual-use offensive capability and potential unauthorized key extraction; use should be restricted to authorized research and isolated test targets.

Confidence: 94%Severity: 76%
AnomalyLOW
guides/hardware-exploitation-patterns-guide.md

This is dual-use hardware-security and penetration-testing guidance. It contains commands capable of unauthorized access, firmware tampering, credential extraction, bootloader bypass, and device damage, but the supplied content does not itself execute those actions or demonstrate malware, persistence, or covert exfiltration. Use only on owned or explicitly authorized equipment and isolated test systems.

Confidence: 98%Severity: 62%
AnomalyLOW
guides/embedded-firmware-analysis.md

This fragment is an offensive-capable hardware security testing guide. It contains explicit reverse-shell, persistence, credential-cracking, authentication-bypass, and firmware-tampering instructions, which are high-risk content if misused. However, the supplied material is static documentation and does not itself perform those actions or show package malware behavior. Further review of the surrounding file would be needed to determine whether these instructions are part of an authorized assessment toolkit or are accompanied by executable automation.

Confidence: 98%Severity: 62%
AnomalyLOW
guides/firmware-extraction-analysis-guide.md

The fragment is readable firmware-analysis and modification guidance. It contains sensitive-data discovery commands and an explicit firmware backdoor/testing and reflashing workflow, creating operational risk if run without authorization. No hidden malicious behavior or supply-chain malware is evident in the shown content; conclusions about omitted portions cannot be made.

Confidence: 96%Severity: 67%
AnomalyLOW
guides/jtag-swd-debugging-guide.md

The fragment is a readable dual-use JTAG/SWD debugging and firmware-extraction guide. It contains no evident malware, credential exfiltration to an external destination, persistence, or host compromise behavior. It does provide actionable instructions for dumping memory and attempting to bypass read protection, creating a significant authorization and confidentiality risk if used without device-owner permission. The guide should be restricted to authorized hardware assessments and controlled lab environments.

Confidence: 98%Severity: 62%
Audit Metadata
Analyzed At
Sep 19, 2026, 04:40 PM
Package URL
pkg:socket/skills-sh/brucesongs%2Fkali-claw%2Fhardware-security%2F@aa45a2cb8667c70d55ed461ca99cca617e3e06a103fc7b79fd2fe7910b34174a