youtube-downloader
Pass
Audited by Gen Agent Trust Hub on May 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/download_video.py) which in turn invokes command-line utilities to perform the video downloads and stream merging. - [EXTERNAL_DOWNLOADS]: The documentation states that the skill automatically installs the
yt-dlppackage from public repositories at runtime if it is not already present on the system. - [DATA_EXFILTRATION]: The skill initiates outbound network connections to YouTube servers to fetch video content and metadata. This is a functional requirement for the skill's purpose but involves communication with an external service.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted metadata from the internet.
- Ingestion points: YouTube video titles, descriptions, and other metadata are processed during the download and naming phase (referenced in
SKILL.md). - Boundary markers: None identified; there are no instructions for the agent to treat external metadata as data rather than instructions.
- Capability inventory: The skill has the ability to write files to
/mnt/user-data/outputs/and execute system commands via the download script. - Sanitization: There is no evidence of sanitization or filtering of the retrieved metadata before it is presented to the agent.
Audit Metadata