grill-design
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses
npxto download and run the@google/design.mdpackage from Google's official NPM registry and GitHub repository. This is a well-known and trusted source. - [COMMAND_EXECUTION]: The agent is instructed to run shell commands using
npxfor linting and exporting the design specification. Evidence:npx @google/design.md lint DESIGN.mdandnpx @google/design.md export .... - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted data from the local repository. 1. Ingestion points: The skill reads
USERS.md,README.md,tailwind.config, CSS files, and product/marketing copy (SKILL.md, step 1). 2. Boundary markers: No explicit instructions are provided to the agent to distinguish between the skill's instructions and potentially malicious instructions embedded in the ingested files. 3. Capability inventory: The skill has the capability to execute shell commands vianpxand write files to the repository (SKILL.md, steps 4 and 5). 4. Sanitization: There is no mention of sanitizing or validating the content read from external files before it is processed.
Audit Metadata