clipcat
Fail
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides installation commands that download scripts from the vendor's website (https://clipcat.ai/cli and https://clipcat.ai/cli.ps1) and pipe them directly into system shells (bash or powershell iex). This is a high-risk pattern that executes remote code without local verification.
- [EXTERNAL_DOWNLOADS]: The skill initiates downloads of CLI binaries and script files from the vendor's domain and interacts with external TikTok data endpoints.
- [COMMAND_EXECUTION]: The skill relies on extensive use of the clipcat CLI for all operations, including sensitive configuration tasks and video processing.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from social media sources (video descriptions, scripts, comments) to generate prompts. Although it suggests using heredocs to isolate prompt content, the underlying data source is attacker-controllable. (Ingestion: TikTok creator posts and video comments; Boundaries: advised heredocs; Capabilities: clipcat CLI execution; Sanitization: heredoc escaping).
Recommendations
- HIGH: Downloads and executes remote code from: https://clipcat.ai/cli - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata