tiangong-kb-esg-search

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses npx to execute the @tiangong-ai/cli package from the NPM registry. As this is the official tool from the skill's author used for its primary function, it is considered a legitimate dependency.
  • [COMMAND_EXECUTION]: Shell commands are executed within a wrapper script that uses bash arrays and jq to parse JSON input. This design ensures that user-provided search queries and parameters are treated as data rather than executable code, preventing command injection.
  • [DATA_EXFILTRATION]: The script includes functionality to load API credentials and service configuration from local .env or user-specified environment files. This is a common and safe pattern for managing sensitive configuration in developer tools.
  • [PROMPT_INJECTION]: While the skill processes user-supplied search queries and external data files, the risk of indirect prompt injection is minimal given the narrow focus on ESG disclosures and the use of a controlled CLI environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 08:21 AM
Security Audit — agent-trust-hub — tiangong-kb-esg-search