arrow-typed-errors

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • PROMPT_INJECTION (SAFE): No attempts to override agent instructions or bypass safety filters were detected. The language is purely instructional and technical.
  • DATA_EXFILTRATION (SAFE): No access to sensitive files or unauthorized network calls. The external links provided are to official documentation for the Arrow library (arrow-kt.io).
  • EXTERNAL_DOWNLOADS (SAFE): No suspicious package installations or remote script executions. The references to external libraries (Quiver, Pedestal State) are conceptual mentions in documentation.
  • COMMAND_EXECUTION (SAFE): No shell commands, privilege escalation, or persistence mechanisms are present. The skill focuses on Kotlin code structure and error modeling.
  • OBFUSCATION (SAFE): No encoded content or hidden characters were found.
  • INDIRECT_PROMPT_INJECTION (SAFE): The skill provides coding patterns and does not process untrusted external data or inputs that could influence the agent's behavior maliciously.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:29 PM