api-tester

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses tools to execute HTTP requests via standard network operations. It explicitly prohibits the execution of shell commands with user input to prevent command injection (CWE-78).
  • [CREDENTIALS_UNSAFE]: The skill implements safe credential management by loading authentication tokens exclusively from an environment variable (API_TEST_TOKEN) rather than hardcoding them, and masks sensitive data in reports.
  • [DATA_EXFILTRATION]: While the skill makes network requests, it includes strict security protocols to mask PII and tokens in logs and reports, and limits requests to HTTPS for external APIs.
  • [PROMPT_INJECTION]: The skill defines clear 'Red Lines' and boundaries to prevent misuse. It does not contain any patterns attempting to override underlying agent safety guidelines.
  • [EXTERNAL_DOWNLOADS]: The skill refers to official GitHub repositories for its source and documentation. These are well-known or trusted sources and do not involve the execution of untrusted remote code.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 01:01 AM
Security Audit — agent-trust-hub — api-tester