security-bounty-hunter

Warn

Audited by Socket on Sep 19, 2026

4 alerts found:

Securityx3Anomaly
SecurityMEDIUM
SKILL.md

SUSPICIOUS: the skill is internally coherent for bug bounty work, but it materially enables an AI agent to perform offensive security testing and exploit validation against real targets. There is no strong evidence of credential harvesting or malware behavior, yet the exploitation-focused capability makes overall security risk high.

Confidence: 89%Severity: 78%
SecurityMEDIUM
payloads.md

The fragment is an offensive security and bug-bounty toolkit containing numerous exploit demonstrations and automation scripts. It presents substantial security and misuse risk because it can scan targets, extract exposed secrets, access cloud metadata, attempt account takeover, execute commands through vulnerable services, and modify cloud functions. The behavior is explicit rather than covert and appears consistent with authorized penetration testing documentation; no clear package-level malware, persistence, or stealthy host compromise is demonstrated in the visible portion. Use should be restricted to explicitly authorized targets.

Confidence: 97%Severity: 90%
AnomalyLOW
guides/attack-surface-discovery-guide.md

The code is a dual-use security reconnaissance guide, not apparent malware. It presents operational and privacy risks because it performs active scanning, may collect credentials or sensitive cloud data, and contains a command-injection flaw in the Python subprocess wrapper due to shell=True and unsanitized interpolation. Use only within explicit scope, avoid printing secrets, and replace shell invocation with argument arrays and strict input validation.

Confidence: 98%Severity: 68%
SecurityMEDIUM
guides/poc-development-guide.md

The fragment is offensive security PoC and training material, not ordinary dependency implementation code. It contains potentially dangerous exploit demonstrations and command-execution/webshell payloads, but no evidence here of autonomous malware or hidden supply-chain behavior. Use should be restricted to authorized testing environments; publication in a general-purpose package would warrant review because the content could facilitate abuse.

Confidence: 98%Severity: 70%
Audit Metadata
Analyzed At
Sep 19, 2026, 04:40 PM
Package URL
pkg:socket/skills-sh/brucesongs%2Fkali-claw%2Fsecurity-bounty-hunter%2F@baa85ab2cef7c0ee8d0bc9aa9dd7a3185b9b3b594bce977f1a014dc704e3f3c5