officecli-word-form

Fail

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides installation commands that pipe remote scripts directly into the system shell. This method poses a risk as it executes code from an external source without prior inspection.\n
  • Evidence (Linux/macOS): curl -fsSL https://d.officecli.ai/install.sh | bash\n
  • Evidence (Windows): irm https://d.officecli.ai/install.ps1 | iex\n- [EXTERNAL_DOWNLOADS]: The skill retrieves installation scripts and references binary releases from external locations hosted on the developer's infrastructure and GitHub.\n
  • Source: https://d.officecli.ai/install.sh\n
  • Source: https://d.officecli.ai/install.ps1\n
  • Source: https://github.com/iOfficeAI/OfficeCLI/releases\n- [INDIRECT_PROMPT_INJECTION]: The skill takes user instructions to define Word form elements, such as content control tags and labels, and interpolates this data into CLI commands for document modification.\n
  • Ingestion points: User-provided form metadata and content processed by the agent to build the document.\n
  • Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are present in the provided templates.\n
  • Capability inventory: The skill can create and modify local .docx files using officecli (add, set, and raw-set operations).\n
  • Sanitization: No explicit sanitization or validation of user-provided content is mentioned in the instructions.
Recommendations
  • HIGH: Downloads and executes remote code from: https://d.officecli.ai/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Oct 5, 2026, 01:40 PM