typography-expert
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is configured with the
WebFetchtool, creating a surface for indirect prompt injection where an agent could ingest malicious instructions from external websites while performing typography analysis. - Ingestion points: Untrusted data can enter the agent's context through the
WebFetchtool when analyzing external URLs or CSS files (SKILL.md). - Boundary markers: There are no explicit delimiters or system instructions defined in the skill to treat fetched web content as untrusted or to ignore embedded instructions.
- Capability inventory: The skill possesses the
Read,Write,Edit, andWebFetchcapabilities. This allows for a potential attack chain where external instructions could influence the agent to modify or overwrite local files. - Sanitization: The skill lacks logic to sanitize, validate, or filter content retrieved via
WebFetchbefore it is processed by the agent.
Audit Metadata