video-lowlight-enhance

Pass

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the meitu command-line utility to perform video processing. User-provided input from the video_url parameter is interpolated directly into the shell command: meitu video-lowlight-enhance --video_url {video_url} --json --download-dir {output_dir}.
  • [EXTERNAL_DOWNLOADS]: The skill includes a preflight check that may install or update the meitu-cli package from the NPM registry using npm install -g meitu-cli@latest. This is a vendor-owned resource used for the skill's primary function.
  • [CREDENTIALS_UNSAFE]: The skill requests permissions to read sensitive authentication data from ~/.meitu/credentials.json and uses environment variables MEITU_OPENAPI_ACCESS_KEY and MEITU_OPENAPI_SECRET_KEY to authenticate with the vendor's service.
  • [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes untrusted external data.
  • Ingestion points: User-provided video_url parameter in SKILL.md.
  • Boundary markers: None present to separate untrusted input from the command execution context.
  • Capability inventory: Execution of shell commands via the meitu CLI tool.
  • Sanitization: No explicit validation or sanitization of the input URL is defined in the skill's workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 15, 2026, 06:03 AM
Security Audit — agent-trust-hub — video-lowlight-enhance