21st-ui-build
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data from
.21st/design.jsonand.21st/DESIGN.mdto influence the agent's UI generation and modification logic. - Ingestion points: Reads configuration and design instructions from the local
.21st/directory. - Boundary markers: The skill does not define clear delimiters or "ignore previous instructions" guards when reading the design context.
- Capability inventory: The agent can execute the
21stCLI for searching, generating, and reviewing code, and has the ability to modify project source files. - Sanitization: No validation or sanitization is mentioned for the content of the design files before they are used to parameterize CLI commands.
- [COMMAND_EXECUTION]: The skill relies on the
21stCLI tool to perform various tasks includinginit,search,generate, andreview. These commands interact with the local filesystem and potentially remote services. - [EXTERNAL_DOWNLOADS]: The workflow involves retrieving or installing UI components from the 21st-dev service/registry to implement interface changes.
Audit Metadata