rightcode-grill-me
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a set of prompting instructions for an interview workflow. It does not perform any network operations, access sensitive credentials, or execute arbitrary code.
- [PROMPT_INJECTION]: The instructions provide behavioral guidelines for the agent (e.g., 'interview me relentlessly'). These are within the intended functional scope and do not attempt to bypass system safety filters or extract administrative prompts. Additionally, while the skill processes user-provided plans and can read the codebase, it does not demonstrate any specific patterns for indirect prompt injection beyond standard capability surface. 1. Ingestion points: User-provided plan or design documentation. 2. Boundary markers: Absent. 3. Capability inventory: Reading local codebase files. 4. Sanitization: Absent.
Audit Metadata