chain-of-verification
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a standard research-based prompting technique (Dhuliawala et al. 2023) designed to improve accuracy and factual grounding.
- [SAFE]: All referenced tools (Read, Grep, Glob, Bash) are standard platform utilities used here for legitimate verification purposes like checking file contents or running tests.
- [SAFE]: There are no signs of prompt injection, data exfiltration, or obfuscation. The instructions focus on reducing errors and verifying security claims during audit phases.
Audit Metadata