primekg

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The skill reads from a specific local file path (/mnt/c/Users/eamon/Documents/Data/PrimeKG/kg.csv). It does not include any network operations or mechanisms to transmit data to external servers.
  • [REMOTE_CODE_EXECUTION]: All logic is implemented using standard data processing libraries for local execution. No dynamic code evaluation or remote script loading was identified.
  • [PROMPT_INJECTION]: No behavioral overrides or safety bypasses were found in the instructions. The skill processes external data from a CSV, which constitutes an indirect prompt injection surface. Ingestion points: The CSV data is loaded via pd.read_csv in scripts/query_primekg.py. Boundary markers: Absent. Capability inventory: Restricted to local data retrieval and pandas manipulation across all scripts. Sanitization: Absent.
  • [COMMAND_EXECUTION]: The skill code does not invoke shell commands, subprocesses, or system-level binaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 10:13 PM