craft
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes project-specific data from files such as
.ui-craft/spec.mdand.ui-craft/brief.md, which creates a vulnerability to Indirect Prompt Injection. \n - Ingestion points: Data is read from
.ui-craft/spec.md(Step 0) and.ui-craft/brief.md(Step 1) to define the surface composition and component inventory. \n - Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore potentially malicious embedded content within these configuration files. \n
- Capability inventory: The skill has the capability to generate and write code (Step 3) and invoke external tools via MCP (
fold_candidates,check_fold). \n - Sanitization: There is no specified logic for sanitizing or validating the contents of the ingested project files before they are processed by the agent context. \n- [COMMAND_EXECUTION]: The skill utilizes external tools and performs build operations as part of its primary function. \n
- Tool Invocation: Executes MCP tools
fold_candidatesandcheck_foldto perform design calculations and visual render checks. \n - Build Process: Orchestrates a multi-step execution sequence to generate software artifacts based on defined build recipes.
Audit Metadata