fill-tsc-jd-lithan-ccp

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads 'Objective' text from user-provided Excel workbooks to serve as instructions for generating job descriptions. This ingestion point in scripts/tsc_jd.py lacks boundary markers or sanitization, potentially allowing malicious workbook content to influence agent output. The skill possesses write capabilities and can execute local scripts, forming a capability chain for processed data.
  • [DYNAMIC_EXECUTION]: The Python script scripts/tsc_jd.py uses subprocess.run to invoke an external script recalc.py located in a relative sibling directory. This runtime execution is used for workbook verification.
  • [COMMAND_EXECUTION]: The skill utilizes the subprocess module to execute local Python commands to handle spreadsheet recalculation tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 07:30 AM
Security Audit — agent-trust-hub — fill-tsc-jd-lithan-ccp