aws-samples/sample-mimir-saga-ai-media-pipelines ? reverse-engineered prompt

Build me this AWS project for broadcast news media workflows.

I want a one place setup that connects Mimir and Saga with AI powered pipelines for video, stories, and rough cuts. It should deploy the cloud pieces, the API, the background jobs, and the MCP server so I can search assets, look up transcripts, create and update stories, and trigger the media actions from tools like Quick Suite or Kiro.

Please make the main flows work end to end, including video summaries, chapter markers, smart vertical reframing, story research, wire feed matching, and rough cut timeline creation. If a video does not already have a transcript, handle that automatically. Use AWS services and the existing deployment setup, and keep it easy to configure with example values and a clear deploy path.

If you need to check current AWS or MCP docs to finish any integration details, go ahead and look them up online.