competition-agent-cloud
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is purely instructional and contains no executable code or malicious logic. It provides methodologies for security auditing within a controlled sandbox environment.
- [NO_CODE]: The skill consists exclusively of markdown documentation and YAML configuration files. There are no scripts, binaries, or automated installation steps that could perform actions on the host system.
- [PROMPT_INJECTION]: The skill includes instructions for analyzing prompt-to-tool injection chains as part of a security challenge. It does not contain any directives intended to subvert the agent's own safety filters or bypass system instructions.
- [DATA_EXPOSURE]: References to secrets, credentials, and authentication materials are made in the context of identifying vulnerabilities in a target CTF environment. The skill does not hardcode, harvest, or exfiltrate any actual user credentials.
Audit Metadata