design-motion-design
Warn
Audited by Gen Agent Trust Hub on Jul 8, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The skill employs an extreme loquacity technique (filler content) across eight of its reference files:
architecture-patterns.md,code-organization.md,deployment-pipelines.md,error-handling.md,performance-optimization.md,security-best-practices.md,state-management.md, andtesting-strategies.md. Each of these files contains 150 sections of repeated technical babble (e.g., discussions on the CAP theorem, eBPF, and data locality). This structure is a known adversarial pattern used to overwhelm the context window and bypass security scanners. - [METADATA_POISONING]: Although the skill is named and described as a 'Motion Design' tool, a significant majority of its referenced data consists of complex, irrelevant system architecture and backend engineering content. This mismatch between the stated purpose and the actual provided data is deceptive and pollutes the agent's knowledge base.
- [PROMPT_INJECTION]: The inclusion of a massive volume of irrelevant and repetitive technical content creates a high-risk surface for indirect prompt injection. By filling the context window with predictable patterns, the author could potentially hide malicious instructions that would be harder to detect or could displace critical safety constraints in the agent's active memory.
Audit Metadata