web-ssrf
Audited by Socket on Sep 19, 2026
4 alerts found:
MalwareAnomalyx2SecurityMALICIOUS. The skill is an offensive exploitation playbook for an AI agent, centered on SSRF-driven secret extraction, internal pivoting, callback-based data leakage, and chained RCE. Its behavior is fundamentally unsafe for an agent skill and far exceeds benign educational or defensive workflow scope.
The visible content is an actionable SSRF exploitation reference and presents substantial dual-use security risk, especially due to cloud credential extraction and internal-network probing examples. It does not itself demonstrate embedded malware or automatic malicious behavior; it is documentation and test commands. Execution against unauthorized targets could cause data exposure or unauthorized access.
This is an explicitly offensive SSRF post-exploitation playbook. It enables internal network reconnaissance, cloud credential theft, secret extraction, arbitrary internal protocol requests, destructive Redis actions, and SSH persistence. It is not itself demonstrated to be malware or an npm package implant, but inclusion in a software supply-chain project would constitute a serious security and policy concern because the examples facilitate unauthorized access and system compromise.
The visible code is an SSRF testing toolkit/documentation fragment. It performs and describes outbound probing, OOB callback monitoring, internal service reachability tests, and protocol-specific payload delivery. These capabilities are dual-use and potentially harmful when used without authorization, but the fragment contains no clear package malware indicators such as credential harvesting, persistence, destructive actions, or covert exfiltration. The main security concern is abuse of the scanner and dangerous escalation examples, including metadata and internal-service probing.