Can I use Antigravity to edit recorded video?

Valmera provides hosted video-editing tools through MCP. Connect the tools, select a recording uploaded to your Valmera account, describe the desired edit and review the preview before final export. The native plugin also supplies workflows for footage editing, shorts and website demos.

Is Valmera listed in the Antigravity Marketplace?

The plugin is available directly from Valmera's public repository. Marketplace inclusion is not confirmed. Use the versioned source and installation steps on this page rather than assuming the plugin is available in the Marketplace.

Does the Gemini CLI extension work unchanged in Antigravity?

The clients use different package and MCP configuration formats. Google offers migration tools, but Valmera also publishes a native Antigravity package. Use that package or the serverUrl configuration on this page; Gemini CLI's httpUrl entry is not the Antigravity CLI configuration shown here.

Edit your video with Antigravity

Bring your recording into the same agent conversation you use to build your product. Connect Antigravity to Valmera, describe the cut you want, review the preview and approve the MP4 for delivery.

Start with a working Antigravity installation and sign-in, Git, and a Valmera account. Upload your source footage in Studio. Uploading is free; hosted editing requires a paid Valmera plan. Antigravity access and usage are separate.

Install the native Valmera plugin

The versioned package includes the hosted editor connection and three skills: footage editing, long-video shorts and website demos. Review the files, then run these commands from a directory where valmera-antigravity does not already exist:

Install Valmera for Antigravity

git clone --branch antigravity-v1.0.0 --depth 1 https://github.com/ABO3SKRALMASAOODI/valmera-mcp.git valmera-antigravity
agy plugin validate --strict ./valmera-antigravity/plugins/antigravity
agy plugin install ./valmera-antigravity/plugins/antigravity
agy plugin list --json

Open /plugin in Antigravity to inspect the installed components. Open /mcp for the bundled valmera_editor connection, complete Valmera browser sign-in when requested, then ask “List my Valmera projects.” Select the project containing your recording.

The plugin is distributed directly from Valmera's repository. Marketplace inclusion is not confirmed. The package was checked with Antigravity CLI 1.5.1; the verification details below distinguish package validation from an authenticated edit.

Connect only the editing tools

For a project-scoped connection without the plugin, merge this into .agents/mcp_config.json in your workspace. Preserve existing servers and other configuration:

Antigravity MCP configuration

{
  "mcpServers": {
    "valmera": {
      "serverUrl": "https://valmera.io/mcp/server"
    }
  }
}

Google documents ~/.gemini/config/mcp_config.json for global configuration. Use /mcp in the CLI to inspect the server; the desktop app exposes MCP controls under Customizations. Choose either this connection or the plugin to avoid duplicate Valmera servers. See Google's MCP instructions for your surface's authentication controls.

Turn a product recording into a release update

Use a recording with a complete demonstration and a visible result. Select its uploaded project, wait for analysis, and start with this brief:

Product-update editing brief

Use Valmera on the uploaded recording I select.
Make a 45-second product update for LinkedIn in 16:9.
Keep the problem, the demonstration and its result in that order.
Remove waiting time and false starts. Keep interface labels readable.
Add accurate captions, keep my voice and do not add music.
Confirm the source and editing cost expectations before starting.
Show a preview; export the version I approve as an MP4.

Keep the story intact. Identify the problem, the action and the result in your source. Ask the agent to name the retained ranges before cutting a long recording.
Make the first pass specific. Set the target duration, framing and caption language. If it needs new media or a website capture, agree on the permitted assets and actions first.
Review as a viewer. Check that the controls remain legible, the narration matches the action and the ending shows the result. Send revision notes with the preview version and output timestamps.
Deliver the approved edit. Export the reviewed version, follow the existing job to completion and open the final MP4. Check its ending, audio and captions before posting it.

The plugin's /edit-video-with-valmera, /turn-long-video-into-shorts and /website-demo-video skills organize those workflows. A local file path is not an upload: use Studio first, or authorize the chosen recording's transfer through the upload workflow. An installed skill alone does not supply source footage or approve billable work.

Moving from Gemini CLI

Antigravity stores workspace MCP entries in .agents/mcp_config.json and uses serverUrl for this remote connection. Gemini CLI settings and its httpUrl field should not be copied unchanged. Google provides an extension import route in its migration guide; the native package above is a separate installation route.

Already have a working Valmera connection? Inspect it before adding another. If you remain on Gemini CLI, use the Gemini-specific guide and its account eligibility notes.

When a step stalls

Plugin not detected: point installation at plugins/antigravity inside the cloned repository, then check agy plugin list --json. The repository root also contains packages for other clients.
Tools need sign-in: inspect the correct MCP entry and finish authentication with the Valmera account containing your footage.
Recording missing: verify the selected account, completed upload and analysis. A successful connection may still have an empty project list.
Export still running: follow its existing job ID. A timeout while observing does not establish that the render failed; avoid duplicate exports.

Sources and verification

On November 6, 2026, a fresh clone of tag antigravity-v1.0.0 passed Antigravity CLI 1.5.1 strict validation: three skills, one MCP server and no findings. The bundled skills matched Valmera's shared source. This checks the package structure; an authenticated Antigravity editing session and rendered output were not tested. The brief above is a proposed workflow.

For a recorded Valmera result, see the cuts-and-crop example. That example is not an Antigravity demonstration.

Frequently Asked Questions

The world’s most advanced AI video editor.

Start with your recording

Upload the footage, choose a project and give your editing agent a clear brief.

Open your video projects

Join thousands of creators.

See pricing →