lov-fill-form
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/fill_form.pyexecutes system commands viasubprocess.runto handle legacy.docfile conversions usingtextutilorLibreOffice. The implementation is secure, utilizing argument lists to prevent shell injection and relying on well-known system utilities. - [PROMPT_INJECTION]: The skill instructions in
SKILL.mddirect the agent to synthesize content for form fields using information extracted from user-provided context documents. This creates a surface for indirect prompt injection if those documents contain malicious instructions. - Ingestion points: Input templates (
--template) and reference context files mentioned in the workflow. - Boundary markers: None identified in the prompt instructions to the agent.
- Capability inventory: File system writes for document output and local command execution for format conversion.
- Sanitization: No specific sanitization or filtering logic is present for the content extracted from context files before it is written to the document.
Audit Metadata