Your agent can edit video.
Chat is a terrible place to review it.

The typed-paragraph problem

Remotion, HyperFrames, and ffmpeg gave coding agents the ability to produce real video. The render is the easy half. Then you watch the output and try to type something like: "around the seven second mark, the title eases in too slowly, and the third emoji in the bottom row overlaps the second one when they bounce." Three sentences, two ambiguities, and the agent still guesses.

Pointing at a frame and saying "this one, make it a crocodile" takes two seconds. That gap is why agentic video editing stalls at the feedback step, not the rendering step.

How the loop works

1. The agent uploads its render to Flask. You get the flask.do link the moment the upload starts, not when processing ends. Open it and watch the video arrive.
2. You review by talking. Hit record, speak, draw on the frame, scrub the timeline. Flask splits your recording into separate comments, each pinned to its timestamp, each with a transcript.
3. The agent reads your feedback as data. Verbatim transcripts, timestamps in seconds, tags, reply threads. It can pull the exact frame you were looking at from its local render to see what you saw.
4. It fixes the cut and uploads a new version. Same asset, same link, v2 on top. You leave another round of feedback or call it done.

While you review, the agent sits on wait_for_feedback — a long-poll that hands it your comments seconds after you leave them. You never paste feedback into a chat window.

What the agent actually receives

Every piece of feedback arrives machine-readable. A voice comment at 0:07 looks like this to the agent:

{
  "type": "recording",
  "timestamp": 7.2,
  "recording": {
    "transcript": "this big emoji here - make it a crocodile,
                   and the bottom row needs seven, not five",
    "ai_summary": "Change hero emoji to crocodile; row of 5 -> 7"
  },
  "tags": [{ "name": "Animation" }],
  "url": "https://flask.do/a/vDbh2?comment=..."
}

Transcripts are sliced per segment, so a two-minute ramble becomes a dozen separate, addressable comments instead of one wall of text.

Connect your agent

Claude Code, one command:

claude mcp add --transport http flask https://api.flask.do/api/mcp/mcp

Or install the plugin, which also teaches the agent the whole review loop:

/plugin marketplace add enritarta/flask-plugin
/plugin install flask@flask

On claude.ai or any other MCP client, add the server URL and paste this into your next session:

Add the Flask MCP server: https://api.flask.do/api/mcp/mcp
(a browser window opens - log in to Flask once)

Then: upload <file> to Flask with upload_file_start and send me the
flask.do link right away, before processing finishes. If I ask for
changes, upload the fix as a new version of the same asset (version_of).

Questions

What Flask doesn't do

Flask reviews video and images. There is no document proofing, no asset management, and uploads cap at 5GB per file. The agent's OAuth sign-in needs a browser once, so fully headless CI runs aren't supported yet. If you need those things, Flask is the wrong tool and we'd rather you find that out here.