brand-guidelines
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to style external artifacts, which serves as an ingestion point for potentially untrusted data containing hidden instructions.
- Ingestion points: Target artifacts for styling as defined in SKILL.md.
- Boundary markers: No delimiters or warnings are provided to prevent the agent from obeying instructions embedded in the artifacts.
- Capability inventory: The skill references 'python-pptx', suggesting the agent may perform file write operations to apply styling.
- Sanitization: There is no mention of sanitizing or validating external input.
- [METADATA_POISONING]: The skill description uses the word 'official' to describe the brand resources it provides. While the data provided is accurate for the Anthropic brand, the skill is authored by a third party, which could be misleading regarding the provenance and official support of the skill.
Audit Metadata