agentleFS
brandonbryant12brandonbryant12/clearframe/skills/gemini-omni/SKILL.md

gemini-omni

Skill1 starsChanged 24 days agoMITView on GitHub

Generate sparse Gemini Omni Flash 1.1 video inserts for a ClearFrame film, using its palette and continuity references while keeping factual graphics and the soundtrack native. Use only when an approved brief needs generated footage.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read docs/continuity.md and the official Omni contract.

    line 8
  2. Distinctive:

    Use an insert only when it adds a useful setting, physical visual or atmosphere.

    line 10

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Gemini Omni inserts

Read docs/continuity.md and the official Omni contract. The model ID is gemini-omni-1.1-flash; scripts/omni.mjs implements the Interactions request/MP4 download contract.

Use an insert only when it adds a useful setting, physical visual or atmosphere. Native text, numbers, charts, captions and source labels remain native. Start with a short shot and aim for <=20% generated runtime unless the brief calls for more.

Declare an asset with kind clip, prompt, seconds, resolution 720p, optional one/two image refs and optional previousInteractionId. Set the root continuity treatment, lighting, camera, movement and motif. The framework adds the palette and shared constraints; reference bytes are hashed for cache invalidation. Two references describe first/last compositions. Keep these within the project and inspect them before generation.

Run plan, then clips DIR --only ID --budget N only within authorized paid scope. Duration is steered in the prompt; the estimate is not a hard billing cap. Output can be shorter or stylistically wrong. Metadata records actual duration, interaction ID and usage. Do not repeat an uncertain paid call automatically.

Review the take, both joins and the full film. Use native video captions/source labels and a continuous shared narration/music mix; the generated clip's audio is excluded. Never use model-generated labels, data, logos or fabricated evidence. check rejects clips too short for the authored duration/offset. Keep the output or choose a shorter beat deliberately.

Choose background (plate.side: full) or half-screen (left|right) before generating. Request enough footage for the beat plus source offset and handles. Never loop a short take: obtain longer coverage or shorten/restructure the beat. plan prints footage problems (take shorter than beat + offset, beat longer than one 10 s take, the same seconds shown twice) and clips refuses to pay for a take that cannot cover its beats. When a beat outruns one take, cut it into different shots rather than reusing one. For business films request a locked camera and restrained subject motion. Inspect native text over moving footage at the start, middle, end and lighting changes, both full size and 360 px; blur is not a contrast guarantee. Use a scrim or opaque panel, or switch to split placement. Record this review before advancing beyond the rough cut. See the B-roll section in docs/continuity.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.