frontend-css
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to refer to an external file path (
../../../agent-os/standards/frontend/css.md) for detailed coding standards, which is a common practice but introduces an indirect prompt injection surface. - Ingestion points:
SKILL.mdreferences a relative path outside the skill directory to fetch instructions. - Boundary markers: There are no delimiters or warnings present to instruct the agent to ignore potentially malicious instructions within the referenced file.
- Capability inventory: The skill is designed for frontend development tasks including modifying
*.css,*.scss, and*.jsfiles, granting the agent file-write capabilities. - Sanitization: No sanitization or validation of the external file's content is specified.
- [EXTERNAL_DOWNLOADS]: The skill's metadata contains an installation command that fetches the skill definition from a repository on GitHub.
- Evidence: The
install_commandinmetadata.jsonusescurlto downloadSKILL.mdfromraw.githubusercontent.com/jeremyrdavis/coffeeshop26/main/.claude/skills/frontend-css/SKILL.md.
Audit Metadata