flaggems-pr-submit-flagos
Pass
Audited by Gen Agent Trust Hub on Jul 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of subprocesses to interact with Git, GitHub CLI (gh), and testing frameworks. Technical review confirms that these commands are executed using list-based arguments without the shell=True flag, effectively preventing command injection vulnerabilities.
- [EXTERNAL_DOWNLOADS]: The skill requires several standard Python libraries (pyyaml, openpyxl, pandas, pytest) and CLI tools (gh, pre-commit). These are reputable dependencies required for its primary development and automation tasks.
- [CREDENTIALS_UNSAFE]: Security analysis of credential handling shows that the skill correctly manages GitHub tokens through environment variables (GH_TOKEN). No hardcoded secrets or unsafe storage patterns were detected in the scripts or instructions.
- [SAFE]: The skill includes advanced security features such as an 'anti-hack' layer in check_operator.py. This layer uses Abstract Syntax Tree (AST) analysis to verify that submitted code does not contain malicious patterns and uses a 'dual-execution' check to ensure that Triton kernels are performing actual computation, protecting the project from deceptive code submissions.
Audit Metadata