data-algo-competitive
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues or malicious patterns were identified during the analysis of the skill and its reference files.
- [PROMPT_INJECTION]: The skill instructions define clear operational phases for the agent (Diagnose, Recommend, Decide, Ship) without attempting to bypass safety filters or override system-level instructions.
- [DATA_EXFILTRATION]: No sensitive file paths, environment variable access, or unauthorized network operations were detected. The skill does not include any tools for data harvesting.
- [REMOTE_CODE_EXECUTION]: There are no patterns involving external script downloads (e.g., curl|bash) or runtime execution of remote content.
- [COMMAND_EXECUTION]: The C++ templates provided are static text intended for competitive programming. The skill does not execute shell commands or request elevated privileges.
- [INDIRECT_PROMPT_INJECTION]: Although the skill processes user-supplied problem statements to identify algorithmic patterns, it lacks any high-risk capabilities (like file writing or network access) that could be exploited through malicious data ingestion.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or secrets were found within the skill metadata or source code templates.
Audit Metadata