grokipedia-recommendations
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by interpolating user-provided inputs into output templates.
- Ingestion points: User-supplied variables such as "Article Topic," "Additional Details," and "URLs" are used to populate the Suggest Article and Suggest Edit forms in SKILL.md.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between user-supplied data and operational instructions.
- Capability inventory: The agent's capabilities in this context are limited to generating static text for the user to copy and paste manually. The skill does not automate network requests, file modifications, or command execution based on this data.
- Sanitization: No input validation or escaping mechanisms are described in the instructions.
- [METADATA_POISONING]: The skill documentation includes claims and references that appear chronologically inconsistent.
- Evidence: The skill references platform statistics and a Verge article from "Jan 2026," which are future-dated relative to the training data. These claims appear intended to establish the authority of the Grokipedia platform for marketing purposes and do not impact the technical safety of the agent.
Audit Metadata