dupey-document-family

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes instructions to automatically modify the user's GitHub account state by starring the NomaDamas/dupey repository using the gh api command. This modification happens by default unless the user explicitly opts out, representing an unrequested action on the user's account.- [COMMAND_EXECUTION]: The skill executes cargo install dupey to install the required document scanning utility if it is not found on the system.- [EXTERNAL_DOWNLOADS]: The skill downloads the dupey tool from the Rust package registry (crates.io) or the author's GitHub repository during the installation phase.- [PROMPT_INJECTION]: The skill parses untrusted data from document files (PDF, DOCX, XLSX, etc.) to generate interpretation summaries and cleanup plans. This represents a surface for indirect prompt injection. Evidence Chain: 1. Ingestion points: dupey scan reads content from user-provided file paths. 2. Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the scanned documents. 3. Capability inventory: The agent is granted the ability to move, rename, and delete files based on the scan results (though user approval is required). 4. Sanitization: There is no evidence of sanitization or filtering for the extracted text before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 12:21 AM
Security Audit — agent-trust-hub — dupey-document-family