The Agent Skills Directory

[SAFE]: The skill provides instructional frameworks for auditing the 'agent-readiness' of public web resources (e.g., mcp.json, openapi.json, llms.txt).\n- [DATA_EXFILTRATION]: The skill describes a process for fetching and analyzing configuration data from remote URLs. This activity is restricted to public-facing metadata intended for discovery and does not involve the exfiltration of sensitive local data or credentials.\n- [PROMPT_INJECTION]: The skill involves ingesting data from external sources, which constitutes an indirect prompt injection surface. However, this is the primary purpose of an auditing tool and no specific patterns intended to subvert agent safety or logic were detected.\n- [REMOTE_CODE_EXECUTION]: While the skill mentions verifying CLI tools and SDK packages, it does so through informational checks (e.g., querying registries or help menus) rather than executing arbitrary remote code or installing untrusted dependencies.

seo-agent