show-me-your-work

Pass

Audited by Gen Agent Trust Hub on Aug 9, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a bash script scripts/log.sh to manage local decision logs in TSV format. The script uses standard system utilities such as mkdir, printf, date, and tr to safely append rows to a file.
  • [DATA_EXFILTRATION]: The skill is designed to store a decision trail locally (decisions.tsv). It includes explicit instructions to the agent to avoid committing secrets, credentials, or private transcript content to the log. Furthermore, it incorporates an unslop requirement and a self-audit phase to verify that no sensitive information is leaked.
  • [SAFE]: The skill demonstrates security awareness by implementing a clean() function in its logging script to mitigate CSV/TSV injection attacks. This function strips newlines and tabs and escapes characters that could be interpreted as formulas in spreadsheet software (e.g., =, +, -, @). No obfuscation, remote code execution, or privilege escalation patterns were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 9, 2026, 03:06 PM
Security Audit — agent-trust-hub — show-me-your-work