theme-factory
Warn
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: There is a discrepancy between the skill's metadata and the provided author context. The SKILL.md frontmatter identifies the author as 'Anthropic' and the source as an 'anthropics' GitHub repository, while the authoritative author context for this skill is 'jetbrains'. This inconsistency can lead to misjudgment of the skill's origin and trust level.\n- [INDIRECT_PROMPT_INJECTION]: The skill includes a 'Create your Own Theme' feature that allows users to provide descriptions to generate custom themes. This represents an input surface where external, untrusted data enters the agent's context.\n
- Ingestion points: User-provided descriptions for custom theme generation in SKILL.md.\n
- Boundary markers: The instructions do not specify any boundary markers or instructions to ignore embedded commands for the generated theme output.\n
- Capability inventory: The skill consists solely of documentation and configuration files; it contains no executable scripts, command line execution capabilities, network tools, or file-writing tools that could be abused through injection.\n
- Sanitization: No evidence of input validation or sanitization is present in the skill instructions.
Audit Metadata