cloud-identity-attack

Warn

Audited by Socket on Sep 19, 2026

7 alerts found:

Securityx5Anomalyx2
SecurityMEDIUM
SKILL.md

SUSPICIOUS/HIGH-RISK. The skill is internally consistent with its stated purpose, but that purpose is to arm an AI agent with offensive cloud-identity attack capability, including credential theft, persistence, and evasion. The markdown examples are not load-time malware, yet the overall skill is dangerous and unsuitable for broad agent deployment because it operationalizes high-impact intrusion workflows against real identity providers.

Confidence: 94%Severity: 91%
SecurityMEDIUM
guides/cloud-identity-attack-entra-id-deep.md

This fragment is offensive cloud-identity tradecraft documentation, not apparent executable malware. It contains detailed instructions capable of credential theft, token and cookie replay, mailbox access, MFA abuse, privilege escalation, and persistence in Microsoft Entra/M365 environments. No covert package behavior or automatic exfiltration mechanism is visible in the supplied text, but the documented commands themselves can facilitate unauthorized access when executed. Review the full file and repository context before inclusion in a software package.

Confidence: 98%Severity: 72%
SecurityMEDIUM
guides/cloud-identity-attack-playbook.md

This fragment is an offensive cloud-identity attack playbook. It does not itself show automatic execution because it is primarily documentation and command examples, but it provides highly actionable procedures for credential theft, phishing, MFA bypass, token abuse, privilege escalation, Golden SAML forgery, persistence, and detection evasion. Treat the surrounding file and package as high risk, especially if these commands are executed automatically or exposed through package scripts. The fragment alone does not establish malware execution, but it strongly indicates harmful offensive capability.

Confidence: 99%Severity: 95%
AnomalyLOW
guides/cloud-identity-attack-okta-auth0-deep.md

This fragment is security-training and penetration-testing documentation containing actionable identity-provider attack and administration examples. It presents significant misuse and credential-exposure risk if copied into unauthorized workflows, especially the JWT/SAML forgery, token replay, API enumeration, user creation, and Action deployment examples. It does not contain executable npm package behavior or clear malware indicators in the shown content. Secrets are represented by placeholders, but real secrets supplied through the commands could be transmitted or printed.

Confidence: 97%Severity: 58%
AnomalyLOW
guides/cloud-identity-attack-privilege-escalation-deep.md

The visible fragment is an offensive cloud identity and privilege-escalation reference with defensive sections. It provides actionable procedures for obtaining tokens, resetting service-principal credentials, pivoting across clouds, and dumping Key Vault secrets, creating substantial misuse risk. However, it contains no directly executable package logic, persistence mechanism, obfuscated payload, or automatic data exfiltration. Malware likelihood is low for this fragment, while the security risk of the documented commands is moderate to high.

Confidence: 94%Severity: 68%
SecurityMEDIUM
guides/cloud-identity-attack-federation-abuse-deep.md

This is an offensive identity-security reference or playbook, not an executable software-supply-chain module. It contains detailed instructions that could enable OAuth consent abuse, credential and token theft, mailbox access, JWT forgery, and federation privilege escalation. The code shown does not itself demonstrate malware or automatic compromise, but incorporating or executing these commands in an untrusted package would create substantial security risk.

Confidence: 98%Severity: 72%
SecurityMEDIUM
guides/cloud-identity-attack-device-code-phishing-deep.md

This is a highly actionable guide to identity phishing and cloud-token abuse, including MFA-mediated token theft, persistence, device registration, and Conditional Access evasion. It presents substantial security risk if used outside an explicitly authorized assessment. However, the supplied artifact is documentation, not executable package code, and it does not itself demonstrate malware execution or data exfiltration.

Confidence: 98%Severity: 86%
Audit Metadata
Analyzed At
Sep 19, 2026, 04:39 PM
Package URL
pkg:socket/skills-sh/brucesongs%2Fkali-claw%2Fcloud-identity-attack%2F@54b5a4513385af5f0492f22d9ada83b6175faa7117c627259c94c75fee1efefb