What problem does it solve? Creating an accurate, project-specific demo video requires manually verifying facts, writing narration, synthesizing voice audio, and rendering video, which is slow and error-prone. This Skill automates that pipeline from VS Code while ensuring every claim in the narration traces back to verified repository evidence. ## Core Features & Use Cases - Evidence-Grounded Narration: Builds a claims ledger from project-understanding reports so every spoken statement cites a repository file, rejecting unsupported claims. - Multiple Production Paths: Produces a zero-install interactive HTML preview with browser speech, a narrated MP4 using Azure Speech or local Piper TTS, or an executive demo with Azure OpenAI scripting and optional Speech Avatar presenter segments. - Governed Approval Workflow: Requires explicit approvals before contacting Azure services, downloading the pinned Piper runtime, installing the FFmpeg renderer, or replacing existing output. - Use Case: After finishing a new API project, invoke the skill to rescan the repository, generate an eight-scene script, audition Azure neural voices, and render a verified two-minute onboarding MP4 under dist/project-video/. ## Quick Start Ask the agent to run /project-video to analyze the current repository and produce a narrated project walkthrough video.
