compress-images

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes Bash tools (cwebp, ls, mkdir, mv) to manage files and perform image compression.
  • [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection via untrusted data.
  • Ingestion points: The agent reads filenames from the target directory and parses local content files to update references.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded instructions in the processed data are present.
  • Capability inventory: The skill allows filesystem manipulation and command execution within the target directory.
  • Sanitization: There is no evidence of sanitization or escaping of filenames or file content before they are used in commands or prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 08:23 AM