gdelt-gkg-fetch
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: A thorough review of the skill's logic and the associated Python script revealed no evidence of malicious intent, obfuscated code, or unauthorized data access. The script follows best practices for tool development, including environment-variable-based configuration and detailed logging.
- [EXTERNAL_DOWNLOADS]: The skill fetches data from GDELT, a well-known and reputable public dataset provider. The script incorporates rigorous validation of downloaded ZIP archives, including integrity checks via CRC, strict UTF-8 decoding validation, and verification of the expected tab-separated column count (defaulting to 27) to prevent processing of malformed data.
- [COMMAND_EXECUTION]: Functionality is implemented through a Python script that relies exclusively on standard library modules. The script is designed to be invoked with specific arguments for filtering and data retrieval and does not employ dangerous functions like eval() or exec() to process external data.
- [PROMPT_INJECTION]: The skill's documentation and prompt templates are strictly functional, focusing on task orchestration and structured JSON output. There are no patterns suggesting attempts to bypass agent safety filters or override core behavioral instructions.
Audit Metadata