add-tts-engine
Warn
Audited by Gen Agent Trust Hub on Apr 19, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructs the agent to clone an external repository from a placeholder URL (
git clone <model-library-url>) and then execute its code (model loading and generation) within a virtual environment. This pattern allows for the execution of untrusted code from arbitrary sources. - [EXTERNAL_DOWNLOADS]: The skill uses
git cloneto fetch external source code based on user-provided input, which could lead to downloading malicious payloads. - [COMMAND_EXECUTION]: The skill utilizes several shell commands such as
mkdir,cd,grep,cat, andjust devto manage files, search through code, and run the application. - [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted data from external repositories (
git clone) and processes it using commands with high capabilities (file writes, project builds). Evidence Chain: 1. Ingestion points: External repositories via git clone. 2. Boundary markers: None. 3. Capability inventory:git clone,cat,mkdir,just dev, file writes to backend and frontend source files. 4. Sanitization: No automated sanitization of external content is specified.
Audit Metadata