The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external files and tool outputs to generate design specifications. While this creates a potential surface for indirect prompt injection, the risk is negligible as the skill lacks dangerous execution capabilities.
Ingestion points: Accesses .moai/project/brand/visual-identity.md and JSON output from a design-copywriting skill.
Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" guards for the ingested content.
Capability inventory: The skill's actions are limited to formatting design tokens (JSON) and providing visual implementation guidelines. No subprocess execution, system file writes, or network exfiltration capabilities are present in the scripts or instructions.
Sanitization: There is no mention of sanitizing or escaping the brand identity content before processing.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local project configuration files and brand identity documents. This behavior is documented and essential for its primary function as a design system expert. No unauthorized network operations or hardcoded credentials were found.
[SAFE]: No patterns of obfuscation, remote code execution, persistence, or privilege escalation were detected. The skill uses standard markdown and YAML frontmatter for its logic.

design-brand-system