query

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches platform-specific binaries and update metadata from the official CDN at agent-cdn.nutrient.io.
  • [REMOTE_CODE_EXECUTION]: Downloads, extracts, and executes a remote binary at runtime to perform document querying tasks. The script handles binary lifecycle management, including platform detection, automated updates, and release ID verification.
  • [COMMAND_EXECUTION]: The bin/query script executes the downloaded platform-specific binary with arguments provided by the agent or user, such as input file paths and search queries.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes and returns content from untrusted documents.
  • Ingestion points: Processes text or Markdown content from external files (e.g., converted PDF outputs) and natural-language strings provided as user queries.
  • Boundary markers: Absent; search results are returned as raw text windows with line numbers, but without delimiting markers or instructions for the agent to ignore embedded commands within that text.
  • Capability inventory: The skill's wrapper script can perform network downloads, modify the local file system (creating directories, symlinks, and state files), and execute binaries.
  • Sanitization: The wrapper script does not perform any validation, escaping, or sanitization of the document content or the query string before it is passed to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 09:23 AM
Security Audit — agent-trust-hub — query