design-system-contribution
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read external files such as
AGENTS.mdand various markdown documents within the repository structure (e.g.,../../../docs/contributors/design/design-system-packages.md). This creates a surface area where instructions embedded in those documents could potentially influence the agent's behavior during a contribution task. - Ingestion points: The agent is directed to read
AGENTS.mdand contribution guides relative to the package location or from the GitHub repository (WordPress/gutenberg). - Boundary markers: The instructions do not provide specific delimiters or warnings for the agent to distinguish between its primary instructions and the content of the external documentation files it reads.
- Capability inventory: The skill facilitates file reading and guided code modification but does not contain direct commands for network exfiltration or dangerous dynamic execution.
- Sanitization: No explicit sanitization or filtering mechanism is defined for the content ingested from the repository's documentation files.
Audit Metadata