Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add manishiitg/coding-agent-loop --skill gemini-omni-videogit clone --depth 1 https://github.com/manishiitg/coding-agent-loopWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/manishiitg/coding-agent-loop/gemini-omni-video)<a href="https://agentmods.dev/skills/manishiitg/coding-agent-loop/gemini-omni-video"><img src="https://agentmods.dev/badge/skills/manishiitg/coding-agent-loop/gemini-omni-video/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/manishiitg/coding-agent-loop/gemini-omni-video"><img src="https://agentmods.dev/badge/skills/manishiitg/coding-agent-loop/gemini-omni-video.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00088 | $0.00930 |
| Opus 5.5 | $0.00035 | $0.00372 |
| Sonnet 5.5 | $0.00018 | $0.00186 |
| Haiku 4.5 | $0.00009 | $0.00093 |
Grade A, and why
gemini-omni-video scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Use Gemini Omni Flash deliberately
Read video-provider-capabilities, fal-ai, video-cinematography, and the
selected endpoint's live fal schema first. Current official starting points:
https://fal.ai/models/google/gemini-omni-flashhttps://fal.ai/models/google/gemini-omni-flash/image-to-videohttps://fal.ai/models/google/gemini-omni-flash/reference-to-video/apihttps://fal.ai/models/google/gemini-omni-flash/edit
This playbook covers the fal-hosted Google model. Use fal-ai, not the direct
google-ai provider skill, for these google/gemini-omni-flash... endpoints.
Resolve and record the exact route, live schema, price, defaults, and limits in
the production's capability record.
Choose the route by intent
- Use
google/gemini-omni-flashto create a new scene from text. - Use
google/gemini-omni-flash/image-to-videoto animate an approved opening composition. - Use
google/gemini-omni-flash/reference-to-videowhen supported references must anchor subjects, objects, style, environment, or other scene details. - Use
google/gemini-omni-flash/editfor an iterative, localized revision of an accepted source video instead of regenerating the whole scene.
The product page describes broad text, image, audio, and video understanding,
but each route can expose a narrower input schema. The current reference route,
for example, documents an ordered image_urls input and zero-based prompt tags
such as <IMAGE_REF_0>. Treat that as a discovery hint: fetch the exact live
schema and use only its fields. Never invent video or audio parameters from the
landing-page description, and never translate zero-based tags into one-based
tokens used by another provider.
Keep references explicit
Create an ordered reference ledger before generation. Record each input's purpose and the exact prompt token that addresses it. Remove redundant or contradictory references. If the live endpoint says references across multiple videos are unsupported, do not work around the restriction by hiding several videos in an undocumented field; select one source, extract purposeful frames, or choose another model.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 79 lines · 88 tokens per session scan A 2bc398065bcd
gemini-omni-video is a skill published in the GitHub repository manishiitg/coding-agent-loop (13 stars, last pushed 1mo ago), licensed MIT. It adds 88 tokens to every session and 930 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
goga-cell-go
Golang rules for implementing CODEMANIFEST contracts.
china-market-localization-strategist
Skill "china-market-localization-strategist" from aimerdoux/wavex-os, covering china market localization strategist, 🧠 your identity & memory, 🎯 your core mission, 1. real-time trend intelligence & signal detection and 2. market opportunity extraction (trend → action).
language-translator
Skill "language-translator" from aimerdoux/wavex-os, covering 🌐 language translator, 🧠 your identity & memory, 🎯 your core mission, 🚨 critical rules you must follow and 📋 your technical deliverables.
korean-business-navigator
Skill "korean-business-navigator" from aimerdoux/wavex-os, covering 🧠 your identity & memory, 💬 your communication style, 🚨 critical rules you must follow, 🎯 your core mission and 📋 your technical deliverables.
seo-specialist
Skill "seo-specialist" from aimerdoux/wavex-os, covering marketing baidu seo specialist, 🧠 your identity & memory, 🎯 your core mission, master baidu's unique search algorithm and build comprehensive china search visibility.
cultural-intelligence-strategist
Skill "cultural-intelligence-strategist" from aimerdoux/wavex-os, covering 🌍 cultural intelligence strategist, 🧠 your identity & memory, 🎯 your core mission, 🚨 critical rules you must follow and 📋 your technical deliverables.