tools

Fail

Audited by Socket on Feb 20, 2026

2 alerts found:

MalwareObfuscated File
MalwareHIGH
package-managers/SKILL.md

[Skill Scanner] URL pointing to executable file detected All findings: [CRITICAL] command_injection: URL pointing to executable file detected (CI010) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: URL pointing to executable file detected (CI010) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] prompt_injection: Detected jailbreak/DAN attempt (PI003) [AITech 1.1] [CRITICAL] command_injection: URL pointing to executable file detected (CI010) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [HIGH] command_injection: Reference to external script with install/setup context (SC005) [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] [HIGH] command_injection: Reference to external script with install/setup context (SC005) [HIGH] command_injection: Reference to external script with install/setup context (SC005) [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] [HIGH] data_exfiltration: Credential file access detected (DE002) [AITech 8.2.3] This document is a benign instructional skill describing system package managers and bootstrap/install examples. It is internally consistent with its stated purpose. The primary security concern is the use of convenience installer patterns that execute remote scripts directly (curl|bash, PowerShell iex) and the recommendation to add third-party repositories without demonstrating integrity verification — these are operational risks rather than evidence of malware. No hardcoded credentials, obfuscated payloads, or explicit exfiltration or backdoor mechanisms are present. LLM verification: The skill fragment is broadly consistent with its stated purpose of describing system-level package managers and setup workflows. However, it includes executable install scripts and external installer URLs within documentation, which modestly elevates risk if an agent were to execute or auto-provision using these blocks without safeguards. The static analysis flags (credential file access, jailbreak phrases, and external script references) indicate potential anomalies but do not prove malicious

Confidence: 95%Severity: 90%
Obfuscated FileHIGH
package-managers/scripts/install.sh

This is a legitimate, convenience bootstrap script that performs discovery, optional updates, and installs official platform package managers. The code itself contains no embedded malicious logic, hardcoded credentials, or obvious exfiltration. However, it uses high-risk patterns: direct execution of remote installer scripts (curl|bash and PowerShell iex) and privileged operations (sudo and elevated PowerShell). These are supply-chain vectors — if the remote endpoints are compromised or TLS is bypassed, arbitrary code execution is possible. Treat invocation as a trust decision; add verification (checksums/signatures, pinning) before executing fetched installers, and avoid running automatically in untrusted automated environments.

Confidence: 98%
Audit Metadata
Analyzed At
Feb 20, 2026, 02:33 AM
Package URL
pkg:socket/skills-sh/Tyler-R-Kendrick%2Fagent-skills%2Ftools%2F@8383d9cb1b9b2211f7d9699dba736352cfeeae6c