thesis-docx
Warn
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local PowerShell scripts (e.g., check_word_com.ps1, normalize_word_styles.ps1) using the -ExecutionPolicy Bypass parameter. This is used to interface with the Microsoft Word COM/DOM API to perform document automation and formatting tasks.\n- [EXTERNAL_DOWNLOADS]: The render_mermaid_figure.ps1 script uses the npx package runner to download and execute the @mermaid-js/mermaid-cli tool from the npm registry. This is used to render Mermaid diagram source code into visual assets for inclusion in the thesis document.\n- [PROMPT_INJECTION]: The skill ingests and parses content from user-provided .docx files during the audit and normalization phases, creating a surface for indirect prompt injection.\n
- Ingestion points: Document content is read and processed in scripts/audit_docx_ooxml.py and scripts/normalize_word_styles.ps1.\n
- Boundary markers: There are no explicit delimiters or instructions to help the agent distinguish between the audited document text and its own instructions.\n
- Capability inventory: The skill has the ability to execute shell commands, run PowerShell scripts, and manipulate local files via Microsoft Word automation.\n
- Sanitization: No sanitization or filtering is applied to the text extracted from the Word documents before it is presented to the agent.
Audit Metadata