create-voiceover-elevenlabs

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted 'user briefs' and 'source assets' to guide the workflow and generate scripts. This represents a surface for indirect prompt injection where malicious input could attempt to override agent instructions.\n
  • Ingestion points: SKILL.md (Workflow Step 1: Read the brief).\n
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for the input brief.\n
  • Capability inventory: Shell execution (curl, jq, ffprobe), file system writes (manifest.json, audio files), and a Python script for JSON parsing.\n
  • Sanitization: No sanitization of the input brief is described.\n- [COMMAND_EXECUTION]: The workflow involves executing shell commands via curl, jq, and ffprobe. Template commands in SKILL.md interpolate environment variables like $VOICE_ID and $VO_TEXT. While the use of jq --arg protects the JSON payload, unsanitized inputs for other variables could pose a command injection risk depending on the agent's execution environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 04:39 PM
Security Audit — agent-trust-hub — create-voiceover-elevenlabs