The Agent Skills Directory

[SAFE]: The skill is a reference guide intended for security auditing and vulnerability research, focusing on SAML/SSO attack vectors.
[COMMAND_EXECUTION]: Contains standard utility commands such as grep, base64, and xmllint. These are used appropriately for processing and formatting data during a manual security analysis workflow.
[DATA_EXPOSURE]: Mentions sensitive file paths like /etc/passwd exclusively as part of an XXE (XML External Entity) payload example for testing purposes. It does not contain code to exfiltrate data or access unauthorized files.
[INDIRECT_PROMPT_INJECTION]: While the skill involves analyzing external SAML assertions which could theoretically contain malicious instructions, the documented process is manual and investigative. There is no evidence of automated processing that would lead to a prompt injection vulnerability.

hunt-saml