form-building
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
pnpm actionCLI command to perform form operations (create-form,update-form). These commands are used to translate user requirements into structured database entries. - [PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks (Category 8) as it processes natural language input from users to generate form structures and SQL drafts.
- Ingestion points: User-provided descriptions and field definitions passed to the
--fieldsand--titlearguments inSKILL.md. - Boundary markers: None identified in the instructional text to separate user input from the command execution context.
- Capability inventory: The skill uses shell commands (
pnpm action) to write data to a SQL database and create URL slugs. - Sanitization: There is no evidence of input validation or sanitization within the provided instructions.
Audit Metadata