pdf-design
Fail
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill instructions direct the agent to read sensitive Google OAuth2 credentials from the path '/home/jamditis/.claude/google/drive-token.json'.
- [DATA_EXFILTRATION]: The skill implements a Python script that transmits local data to external Google Drive folders (e.g., '1lKTdwq4_5uErj-tBN112WCdJGD2YtetO') using retrieved authentication tokens.
- [COMMAND_EXECUTION]: The skill executes multiple shell commands, including 'chromium-browser' in headless mode, 'pdftoppm', and 'pdfinfo', and runs a Python script via a shell heredoc.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted HTML data ('new-report.html') and processes it within automated workflows involving shell execution and network operations without boundary markers or sanitization.
Recommendations
- AI detected serious security threats
Audit Metadata