rfc-specification
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [SAFE] (SAFE): No malicious patterns, prompt injection attempts, or obfuscated content were identified in the provided documentation.
- [NO_CODE] (LOW): The skill consists entirely of Markdown templates and guidelines. There are no executable scripts or package requirements, which significantly limits the attack surface.
- [Indirect Prompt Injection] (INFO): The skill defines a surface for processing untrusted technical inputs. 1. Ingestion points: Technical descriptions and stakeholder feedback processed during RFC drafting. 2. Boundary markers: The template utilizes structured Markdown headers and includes a dedicated 'Security Considerations' section. 3. Capability inventory: None; the skill is limited to text generation and does not include executable scripts or system-level tools. 4. Sanitization: No explicit sanitization or input validation logic is present in the markdown templates.
Audit Metadata