invideo
Fail
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill recommends installing its CLI using a shell script from the vendor's GitHub repository piped directly to the shell (
curl -fsSL https://raw.githubusercontent.com/invideoio/cli/main/install.sh | sh). This execution of unverified remote code bypasses package integrity checks and executes with user-level shell permissions. - [EXTERNAL_DOWNLOADS]: The skill fetches installation scripts and configuration from
github.com/invideoioand references documentation atdocs.invideo.io. These are vendor-owned resources. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted media and project data ingested through
/invideo-mediaor user-provided URLs. This surface could be exploited if malicious instructions are embedded in media transcripts or project metadata. Mandatory Evidence: Ingestion via media paths and project URLs; Boundary markers are absent; Capability inventory includes file system writes and network access via CLI tools; Sanitization is not explicitly documented.
Recommendations
- HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/invideoio/cli/main/install.sh - DO NOT USE without thorough review
Audit Metadata