Ai Gemini Video Editor

Drop a video and describe what you want — and watch Gemini's multimodal intelligence turn your raw footage into polished content. The ai-gemini-video-editor...

Security Scan

The skill's runtime instructions broadly match a cloud-based video editor, but metadata mismatches, filesystem checks, hidden token handling, and lack of privacy/hosting clarity are inconsistent or unexplained — review before installing.

Purpose & Capability

The skill claims to provide Gemini-powered video editing and requires a NEMO_TOKEN for API access — that credential is coherent with a cloud backend. However, the registry metadata earlier listed no config paths while the SKILL.md frontmatter declares a config path (~/.config/nemovideo/) — this is an internal inconsistency. Also the name references Google 'Gemini' but all runtime endpoints point to mega-api-prod.nemovideo.ai (not an obvious Google domain), which could be misleading to users.

Instruction Scope

Runtime instructions direct the agent to: obtain an anonymous token automatically (POST to an external endpoint), create and store session IDs, upload user video files/URLs to the remote backend, and detect install platform by checking user paths (e.g., ~/.clawhub/, ~/.cursor/skills/). Uploading full video content to an external service is central to functionality but the skill does not include an explicit privacy/consent notice or explain where session state/tokens are stored. The instruction to 'Don't display raw API responses or token values to the user' increases opacity about credentials and responses.

Install Mechanism

No install spec and no code files — this is an instruction-only skill. That minimizes disk-write/install risk; there is no package download or archive extraction to review.

Credentials

Only one environment credential is declared (NEMO_TOKEN), which is proportionate for a cloud API. However, the SKILL.md both treats NEMO_TOKEN as required and also instructs the agent to auto-request an anonymous token if none is set — that is inconsistent. The frontmatter also lists a config path that the registry metadata did not. There are no other unrelated secrets requested.

Persistence & Privilege

always is false and the skill is user-invocable, so it has normal invocation privileges. The instructions expect the agent to store a session_id and potentially use ~/.config/nemovideo/ for state; the skill does not detail where or how long session info is saved. Reading install paths to set an X-Skill-Platform header is minor but does access user filesystem state.

What to consider before installing

This skill will upload your videos and related data to https://mega-api-prod.nemovideo.ai and will either use a provided NEMO_TOKEN or generate an anonymous token on your behalf. Before installing: (1) verify who runs the nemovideo.ai backend (privacy, retention, and terms); (2) ask where session data and tokens are stored (the frontmatter mentions ~/.config/nemovideo/ but registry metadata omitted it); (3) avoid sending sensitive/private footage until you're comfortable with the service; (4) prefer supplying a token you control rather than relying on an opaque anonymous token flow; and (5) request a homepage/source repository for the skill to inspect implementation and privacy docs. The inconsistencies above (metadata vs. SKILL.md, and the 'Gemini' branding vs. nemovideo.ai endpoints) are reasons to pause.

Getting Started

Welcome to the AI Gemini Video Editor — your intelligent editing assistant that understands your footage and turns your ideas into polished video content. Drop your video or describe your project and let's start editing!

Try saying:

• "I have a 12-minute product walkthrough video. Can you identify the 5 most important moments and suggest where to cut it down to under 3 minutes for social media?"
• "Generate accurate subtitles for this video and reformat the captions to fit a vertical 9:16 TikTok layout with bold on-screen text styling."
• "Analyze this talking-head interview and rewrite the spoken script into a tighter narrative — remove filler words, redundant sections, and suggest a stronger opening hook."

First-Time Connection

When a user first opens this skill, connect to the processing backend automatically. Briefly let them know (e.g. "Setting up...").

Authentication: Check if NEMO_TOKEN is set in the environment. If it is, skip to step 2.

1. Obtain a free token: Generate a random UUID as client identifier. POST to https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token with header X-Client-Id set to that UUID. The response data.token is your NEMO_TOKEN — 100 free credits, valid 7 days.
2. Create a session: POST to https://mega-api-prod.nemovideo.ai/api/tasks/me/with-session/nemo_agent with Authorization: Bearer <token>, Content-Type: application/json, and body {"task_name":"project","language":"<detected>"}. Store the returned session_id for all subsequent requests.

Keep setup communication brief. Don't display raw API responses or token values to the user.

Edit Smarter, Not Harder With Gemini AI

The AI Gemini Video Editor brings the power of Google's Gemini multimodal model directly into your editing workflow. Instead of manually scrubbing through timelines, you describe what you want — and the skill interprets your footage, identifies key moments, and delivers actionable edits, structured scripts, and scene-level suggestions.

Whether you're cutting a 45-minute interview down to a punchy 3-minute highlight reel, adding context-aware subtitles to a product demo, or restructuring a tutorial for better audience retention, this skill understands the content of your video — not just its metadata. It reads scenes, spoken words, visual cues, and pacing to give you edits that actually make sense.

This is built for solo creators, small marketing teams, online educators, and social media managers who produce video regularly but don't have the time or budget for professional post-production. The AI Gemini Video Editor shortens the gap between raw footage and publish-ready content dramatically.

Gemini Routing Your Edit Requests

Every prompt you send — whether trimming a clip, generating captions, or applying a scene transition — gets parsed by Gemini's multimodal understanding layer and routed to the appropriate video processing pipeline automatically.

User says...	Action	Skip SSE?
"export" / "导出" / "download" / "send me the video"	→ §3.5 Export	✅
"credits" / "积分" / "balance" / "余额"	→ §3.3 Credits	✅
"status" / "状态" / "show tracks"	→ §3.4 State	✅
"upload" / "上传" / user sends file	→ §3.2 Upload	✅
Everything else (generate, edit, add BGM…)	→ §3.1 SSE	❌

Cloud API Reference Guide

AI Gemini Video Editor offloads all transcoding, frame analysis, and generative editing tasks to Google's cloud backend, meaning your local machine handles only the interface while Gemini processes the heavy video workloads remotely. API calls are authenticated per session and throttled based on your active credit tier.

Skill attribution — read from this file's YAML frontmatter at runtime:

• X-Skill-Source: ai-gemini-video-editor
• X-Skill-Version: from frontmatter version
• X-Skill-Platform: detect from install path (~/.clawhub/ → clawhub, ~/.cursor/skills/ → cursor, else unknown)

All requests must include: Authorization: Bearer <NEMO_TOKEN>, X-Skill-Source, X-Skill-Version, X-Skill-Platform. Missing attribution headers will cause export to fail with 402.

API base: https://mega-api-prod.nemovideo.ai

Create session: POST /api/tasks/me/with-session/nemo_agent — body {"task_name":"project","language":"<lang>"} — returns task_id, session_id.

Send message (SSE): POST /run_sse — body {"app_name":"nemo_agent","user_id":"me","session_id":"<sid>","new_message":{"parts":[{"text":"<msg>"}]}} with Accept: text/event-stream. Max timeout: 15 minutes.

Upload: POST /api/upload-video/nemo_agent/me/<sid> — file: multipart -F "files=@/path", or URL: {"urls":["<url>"],"source_type":"url"}

Credits: GET /api/credits/balance/simple — returns available, frozen, total

Session state: GET /api/state/nemo_agent/me/<sid>/latest — key fields: data.state.draft, data.state.video_infos, data.state.generated_media

Export (free, no credits): POST /api/render/proxy/lambda — body {"id":"render_<ts>","sessionId":"<sid>","draft":<json>,"output":{"format":"mp4","quality":"high"}}. Poll GET /api/render/proxy/lambda/<id> every 30s until status = completed. Download URL at output.url.

Supported formats: mp4, mov, avi, webm, mkv, jpg, png, gif, webp, mp3, wav, m4a, aac.

SSE Event Handling

Event	Action
Text response	Apply GUI translation (§4), present to user
Tool call/result	Process internally, don't forward
heartbeat / empty data:	Keep waiting. Every 2 min: "⏳ Still working..."
Stream closes	Process final response

~30% of editing operations return no text in the SSE stream. When this happens: poll session state to verify the edit was applied, then summarize changes to the user.

Backend Response Translation

The backend assumes a GUI exists. Translate these into API actions:

Backend says	You do
"click [button]" / "点击"	Execute via API
"open [panel]" / "打开"	Query session state
"drag/drop" / "拖拽"	Send edit via SSE
"preview in timeline"	Show track summary
"Export button" / "导出"	Execute export workflow

Draft field mapping: t=tracks, tt=track type (0=video, 1=audio, 7=text), sg=segments, d=duration(ms), m=metadata.

Timeline (3 tracks): 1. Video: city timelapse (0-10s) 2. BGM: Lo-fi (0-10s, 35%) 3. Title: "Urban Dreams" (0-3s)

Error Handling

Code	Meaning	Action
0	Success	Continue
1001	Bad/expired token	Re-auth via anonymous-token (tokens expire after 7 days)
1002	Session not found	New session §3.0
2001	No credits	Anonymous: show registration URL with ?bind=<id> (get <id> from create-session or state response when needed). Registered: "Top up credits in your account"
4001	Unsupported file	Show supported formats
4002	File too large	Suggest compress/trim
400	Missing X-Client-Id	Generate Client-Id and retry (see §1)
402	Free plan export blocked	Subscription tier issue, NOT credits. "Register or upgrade your plan to unlock export."
429	Rate limit (1 token/client/7 days)	Retry in 30s once

Quick Start Guide

Getting started with the AI Gemini Video Editor is straightforward. Begin by sharing your video file directly in the chat, or paste a video URL if your content is hosted online. Then describe your editing goal in plain language — for example, 'cut this to 90 seconds for Instagram' or 'add chapter markers based on topic changes.'

Gemini will analyze the video's visual content, spoken audio, and scene structure before responding with specific edit recommendations, generated scripts, caption drafts, or a restructured timeline outline. You can then refine the output conversationally — ask for a shorter version, a different tone, or alternative cut points.

For best results, be specific about your target platform, audience, and desired length upfront. The more context you provide, the more precise and useful the editing output will be. You can also ask the skill to explain why it made certain suggestions, which is especially useful for learning better editing instincts over time.

Performance Notes

The AI Gemini Video Editor performs best with videos that have clear audio and reasonably stable footage. Heavily compressed files or videos with significant background noise may result in less precise transcript-based edits, so higher-quality source files will always yield sharper recommendations.

For longer videos — anything over 20 minutes — consider breaking the footage into logical segments before submitting. This helps Gemini focus its analysis and produce more granular, scene-specific suggestions rather than broad structural notes. If you're working with multi-camera footage or a rough cut with rough transitions, mention that context explicitly so the skill can tailor its recommendations accordingly.

Gemini's multimodal understanding means it can process both what is said and what is shown simultaneously, which gives it an edge on content like tutorials, product reviews, and interviews where visual and verbal information need to align. Expect the most detailed outputs for content-dense, dialogue-driven videos.