vercel-react-native-skills
Warn
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [METADATA_POISONING]: The YAML frontmatter in
SKILL.mdidentifies the author as 'vercel', which contradicts the distribution context identifying the author as 'ma1orek'. This mismatch is misleading regarding the skill's origin.- [INDIRECT_PROMPT_INJECTION]: The skill provides extensive instructional content across 38 files designed to override and shape agent behavior during mobile development tasks. - Ingestion points: The entire skill structure, including
SKILL.md,AGENTS.md, and therules/directory. - Boundary markers: None. The instructions are direct and intended to be adopted as primary guidelines.
- Capability inventory: No specific tools or scripts are shipped with the skill; it leverages the agent's base coding capabilities.
- Sanitization: Not applicable to static documentation content.
Audit Metadata