cs-methodology-evaluation
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is comprised of markdown instructions and YAML metadata designed to guide research auditing. It does not contain any executable scripts, shell commands, or binaries.\n- [NO_CODE]: The absence of functional code (Python, Node.js, Shell) precludes risks such as command injection, remote code execution, or unauthorized file system access.\n- [SAFE]: The
references/repository-contract.mdfile explicitly promotes data security by instructing that generated configuration files must never contain sensitive information like API keys, tokens, or session data.\n- [SAFE]: No patterns of prompt injection, obfuscation, or data exfiltration were detected. The skill is designed to improve scientific integrity rather than override agent security protocols.
Audit Metadata