genmedia

Fail

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructs the user or agent to install the genmedia CLI by piping a remote script to bash via 'curl https://genmedia.sh/install -fsS | bash'.
  • [DYNAMIC_EXECUTION]: Windows installation instructions use 'Invoke-Expression' (iex) to execute a remote PowerShell script from 'https://genmedia.sh/install.ps1'.
  • [EXTERNAL_DOWNLOADS]: The skill performs multiple external operations including downloading installation scripts, fetching skill bundles with 'genmedia init', and installing new skills from a registry with 'genmedia skills install'.
  • [COMMAND_EXECUTION]: The skill operates by running shell commands through the 'genmedia' CLI to interact with model endpoints and manage files.
  • [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection. (1) Ingestion points: User prompts are passed to 'genmedia run' and 'genmedia models'. (2) Boundary markers: No delimiters are specified to isolate data. (3) Capability inventory: The skill can execute commands, upload/download files, and access network endpoints. (4) Sanitization: No input validation is present.
Recommendations
  • HIGH: Downloads and executes remote code from: https://genmedia.sh/install - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 14, 2026, 04:39 PM
Security Audit — agent-trust-hub — genmedia