webcraft-init
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from existing workspace files to automate project setup, creating a surface for indirect prompt injection from potentially untrusted project data. \n
- Ingestion points: The skill reads
00-project.md,package.json,components.json, and other project-related files to identify framework, tooling, and design system details (found inSKILL.md). \n - Boundary markers: There are no instructions in the skill defining clear delimiters or warnings to ignore potentially malicious instructions embedded within the ingested project files. \n
- Capability inventory: The skill is capable of writing new project records (
00-project.md) and establishing project directories based on the findings of its analysis. \n - Sanitization: The instructions do not specify any validation, escaping, or filtering for data extracted from the existing environment before it is used to influence the agent's output.
Audit Metadata