gpd-submission-health

Pass

Audited by Gen Agent Trust Hub on Feb 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONNO_CODE
Full Analysis
  • COMMAND_EXECUTION (SAFE): The skill lists several bash commands intended for the gpd CLI. These commands perform standard developer tasks such as validating releases, checking status, and uploading deobfuscation files. There are no signs of arbitrary or malicious command execution.
  • DATA_EXFILTRATION (SAFE): No sensitive file paths (e.g., SSH keys, credentials) are accessed. The skill only interacts with application-specific metadata and assets like mapping.txt and icon.png for publishing purposes.
  • NO_CODE (SAFE): The skill consists entirely of markdown documentation and shell examples. It does not include or download external scripts or binary executables.
  • PROMPT_INJECTION (SAFE): No patterns of instruction override, jailbreak attempts, or safety filter bypasses were detected in the markdown content or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 23, 2026, 03:08 AM