mols-documents-studio
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a documentation manager using predefined templates and workflows. It does not contain remote code execution, credentials, or persistence mechanisms.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (Category, Title, Decision Detail, Reason Detail, Impact Detail) and interpolates them directly into markdown templates (
workflows/write-decisions.md). While this is intended functionality, the lack of boundary markers or sanitization could allow malicious content in these fields to influence subsequent agent actions if those documents are later read by an AI. - Ingestion points: User-provided parameters in
workflows/write-decisions.md(Target Path, Category, Title, Decision Detail, Reason Detail, Impact Detail). - Boundary markers: None present; the instructions do not specify delimiters for user-provided strings or warnings to ignore embedded instructions.
- Capability inventory: File read/write operations for documentation management (SKILL.md, workflows/write-decisions.md).
- Sanitization: No explicit sanitization or validation of input fields is performed beyond length constraints on the Title field.
Audit Metadata