officecli
Fail
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill contains instructions to download and execute scripts from 'https://d.officecli.ai/install.sh' via bash and 'https://d.officecli.ai/install.ps1' via PowerShell's iex. This provides a mechanism for arbitrary remote code execution from an unverified domain.
- [COMMAND_EXECUTION]: The skill's primary functionality relies on executing shell commands through the 'officecli' tool to create, modify, and analyze Office files.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external Office files, representing an attack surface for indirect prompt injection. 1. Ingestion points: 'officecli view', 'officecli get', and 'officecli query' in SKILL.md. 2. Boundary markers: None identified in instructions. 3. Capability inventory: File system modifications (add, set, remove, batch), local server execution (watch), and dynamic plugin loading. 4. Sanitization: No evidence of data sanitization or boundary enforcement before processing content.
Recommendations
- HIGH: Downloads and executes remote code from: https://d.officecli.ai/install.sh - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata