interactive-explainer
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes
ffmpegfor concatenating video clips andffprobefor checking the duration of audio files. These are standard utility calls necessary for the skill's primary function of media assembly. - [EXTERNAL_DOWNLOADS]: The skill references several external sub-skills hosted on the
PrunaAIGitHub organization (e.g.,p-image,p-video-avatar) and providesnpxcommands for their installation. These resources belong to the same vendor as the analyzed skill and are documented neutrally. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied content such as video topics and character dialogue.
- Ingestion points: Data enters through the intake process described in
SKILL.mdand is structured intemplates/explainer-plan.template.json. - Boundary markers: The skill uses a structured JSON template to separate project components.
- Capability inventory: The agent executes shell commands (
ffmpeg,ffprobe,npx) and interacts with several generative video and audio APIs. - Sanitization: The skill provides detailed 'positive prompt' rules and 'blocked phrase' lists in
references/interactive-explainer-prompts.mdto prevent model confusion and poor output quality.
Audit Metadata