hve-spielberg

Fail

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: Provides instructions to download and execute a requirements check script from the author's repository. The script content is included in the skill for inspection.
  • [REMOTE_CODE_EXECUTION]: Python scripts automatically attempt to install the 'requests' library via the system package manager if it is missing.
  • [COMMAND_EXECUTION]: Extensively uses subprocess calls to execute media processing tools such as ffmpeg, ffprobe, and asciinema.
  • [EXTERNAL_DOWNLOADS]: Fetches configuration and media assets from established services including ElevenLabs, Freesound, and the jsDelivr CDN.
  • [PROMPT_INJECTION]: Implements a capture workflow that uses project descriptions and metadata to generate shell commands for terminal recording, which represents an indirect prompt injection surface.
  • [PROMPT_INJECTION]: Includes a mock review file (REVIEW.md) with self-referential security claims that are treated as informational data and do not influence the final verdict.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/nebrass/hve-spielberg/main/scripts/check_requirements.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 24, 2026, 04:46 PM
Security Audit — agent-trust-hub — hve-spielberg