office-oxide-mcp-rust-document-processor
Warn
Audited by Gen Agent Trust Hub on Jun 15, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to download and install binary files or source code from an unverified GitHub organization ("Aimino-Tech") via
cargo install,git clone, and direct downloads from the releases page. - [REMOTE_CODE_EXECUTION]: Instructs the user to build and run code from a remote, unverified source using
cargo installandcargo build, which can execute arbitrary code during the compilation and installation process. - [COMMAND_EXECUTION]: Provides shell commands for software installation and configuration, including
cargo install,git clone, and environment modifications for MCP (Model Context Protocol) clients. - [PROMPT_INJECTION]: Detects a surface for indirect prompt injection because the skill ingests and processes complex untrusted data formats (XLSX, DOCX, PPTX, PDF) that may contain hidden malicious instructions.
- Ingestion points: The
office_readtool in SKILL.md reads content from external documents. - Boundary markers: No explicit boundary markers or warnings to ignore embedded instructions were found.
- Capability inventory: The skill possesses extensive local file system capabilities, including writing to spreadsheets (
office_write_cell), creating new documents (office_create_docx), filling PDF forms (office_fill_pdf_form), and inserting images (office_add_image). - Sanitization: No evidence of sanitization or validation of the content extracted from external documents before it is processed by the AI.
Audit Metadata