← 返回 Skills 市场
peand-rover

Ai Video Color Grading

作者 peandrover adam · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
137
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install ai-video-color-grading
功能描述
Professional color correction and color grading for any video with AI — transform flat raw footage into cinematic, branded, or mood-specific visuals. NemoVid...
使用说明 (SKILL.md)

AI Video Color Grading — Make Every Frame Look Like It Was Shot by a Cinematographer

Color grading is the invisible art that separates amateur video from professional production. The same footage — identical composition, identical lighting, identical subject — looks completely different after color grading. Warm tones communicate comfort, nostalgia, and trust. Cool tones communicate technology, precision, and modernity. Desaturated tones communicate drama, seriousness, and weight. High-contrast grading communicates energy and boldness. The color grade is the emotional filter through which the viewer experiences every frame. Professional colorists charge $100-500 per hour. A 10-minute video takes 2-4 hours to grade manually in DaVinci Resolve. A multi-camera project requires matching colors across sources — adding hours of node-by-node adjustment. Consistency across a content library (every video matching the same look) requires saving and maintaining custom LUTs, adjusting for different lighting conditions, and reviewing every export. NemoVideo grades video through natural language. Describe the look you want — "warm cinematic like a Wes Anderson film" or "clean professional corporate" or "moody blue-toned thriller" — and the AI applies: technical correction (exposure, white balance, noise reduction) followed by creative grading (color temperature, contrast curve, saturation mapping, highlight/shadow tinting, and skin tone protection). Professional color grading from a description.

Use Cases

  1. YouTube Creator — Signature Look Across All Videos (any length) — A creator shoots on different cameras, in different locations, with different lighting — every video looks slightly different. NemoVideo: establishes a signature grade ("warm-clean: slight warm tone, lifted shadows, gentle contrast, natural skin preservation"), applies it consistently to every video regardless of source camera or lighting conditions, and maintains the look across an entire channel library. Viewers develop subconscious brand recognition — the color grade becomes part of the creator's visual identity.

  2. Multi-Camera Match — Wedding/Event Videography (multiple sources) — A wedding was shot on 3 cameras: a Sony (neutral color science), a Canon (warm skin tones), and a GoPro (action cam with aggressive processing). Cut together without grading, the video jumps between three different color worlds at every camera switch. NemoVideo: analyzes the color characteristics of each source, creates a unified grade that brings all three cameras into the same visual space (matched white balance, matched exposure, matched contrast), preserves the best qualities of each (Canon's skin tones applied to all three), and exports with invisible camera transitions. Three cameras look like one.

  3. Cinematic Grade — Film Look for Any Footage (any length) — A filmmaker shoots on a consumer camera but wants a cinematic look. NemoVideo applies: lifted blacks (the signature "film" look — shadows never reach true black), subtle halation on highlights (the glow that film grain creates on bright areas), reduced saturation with specific color shifts (teal shadows, orange highlights — the most popular cinematic color combination), gentle film grain overlay (texture that adds organic feeling), and controlled contrast curve (mid-tones compressed for that "filmic" latitude). Phone footage or consumer camera footage with the visual character of high-end cinema cameras.

  4. Brand Consistency — Corporate Video Library (batch) — A corporation has 50 training and marketing videos produced over 3 years by different teams. Some are warm, some cold, some overexposed, some dark. NemoVideo batch-grades the entire library: normalizes exposure and white balance across all 50 videos, applies the brand's visual standard (specific color temperature, contrast level, saturation range), ensures skin tones look healthy and consistent, and exports the entire graded library. Fifty videos that previously looked like they came from fifty different companies now look like one cohesive collection.

  5. Mood Transformation — Same Scene, Different Feeling (any length) — A real estate video was shot on an overcast day. The listing looks gray and uninviting. NemoVideo: warms the color temperature (makes overcast feel like golden hour), lifts brightness (makes rooms feel airier and more spacious), enhances green in outdoor shots (makes landscaping look lush), brightens windows (makes natural light feel abundant), and applies a "warm inviting home" grade. The same listing that looked depressing in raw footage now looks like a dream home. Color grading does not change what was filmed — it changes how it feels.

How It Works

Step 1 — Upload Video

Any footage: phone, consumer camera, professional camera, screen recording, drone, action cam. NemoVideo handles any source quality.

Step 2 — Describe the Look

Natural language: "warm and inviting like a coffee shop commercial" or "cold and dramatic like a thriller" or "clean corporate with brand blue tones." Or choose a preset: cinematic, warm-clean, cool-modern, vintage-film, vibrant-social.

Step 3 — Generate

curl -X POST https://mega-api-prod.nemovideo.ai/api/v1/generate \
  -H "Authorization: Bearer $NEMO_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "skill": "ai-video-color-grading",
    "prompt": "Apply cinematic color grading to a 10-minute short film. Style: warm cinematic with teal shadows and orange highlights. Lifted blacks (no true black — darkest value around 10%%). Subtle film grain (fine, not distracting). Slightly desaturated overall but with enhanced skin warmth. Highlight roll-off should feel smooth and organic (not digital clipping). Maintain consistent grade across all scenes (indoor and outdoor). Preserve natural skin tones even in stylized scenes.",
    "grade_style": "cinematic-teal-orange",
    "corrections": {
      "exposure": "auto-normalize",
      "white_balance": "auto-correct",
      "noise_reduction": "subtle"
    },
    "creative": {
      "shadows": "teal-shifted",
      "highlights": "warm-orange",
      "blacks": "lifted-10pct",
      "saturation": "reduced-15pct",
      "skin_tones": "protected-warm",
      "grain": "fine-subtle",
      "highlight_rolloff": "smooth-organic"
    },
    "consistency": "match-across-scenes",
    "format": "16:9"
  }'

Step 4 — Review and Refine

Preview the graded footage. Compare before/after. Adjust: "make the shadows more blue," "increase the warmth slightly," "reduce the grain." Each refinement applies instantly.

Parameters

Parameter Type Required Description
prompt string Describe the visual look you want
grade_style string "cinematic-teal-orange", "warm-clean", "cool-modern", "vintage-film", "vibrant-social", "moody-dark", "corporate-clean"
corrections object {exposure, white_balance, noise_reduction} — technical fixes
creative object {shadows, highlights, blacks, saturation, skin_tones, grain, contrast}
consistency string "match-across-scenes", "match-to-reference", "per-scene-auto"
reference string Reference image or film name to match
skin_tones string "protected" (default), "stylized", "warm", "cool"
batch boolean Grade multiple videos with same look
export_lut boolean Export the grade as a reusable LUT file
format string "16:9", "9:16", "1:1"

Output Example

{
  "job_id": "avcg-20260328-001",
  "status": "completed",
  "source_duration": "10:22",
  "grade_applied": "cinematic-teal-orange",
  "corrections": {
    "exposure_adjusted": "3 scenes normalized",
    "white_balance_corrected": "2 scenes (indoor tungsten → neutral)",
    "noise_reduction": "applied to 1 low-light scene"
  },
  "creative_grade": {
    "shadows": "teal-shifted (#1A3A4A)",
    "highlights": "warm-orange (#FFB088)",
    "blacks_lifted": "10%%",
    "saturation": "-15%% overall, skin +5%%",
    "grain": "fine (opacity 8%%)"
  },
  "consistency": "matched across 12 scenes",
  "output": {
    "file": "short-film-graded.mp4",
    "resolution": "1920x1080",
    "before_after_preview": "comparison-strip.png"
  }
}

Tips

  1. Correction before creativity — Always fix technical issues first (exposure, white balance, noise) before applying creative grading. A warm cinematic grade on underexposed footage with wrong white balance produces muddy results. Clean footage grades beautifully.
  2. Skin tone protection is non-negotiable — Creative grading that makes skin look green, gray, or unnaturally saturated immediately looks amateur. Always enable skin tone protection — it allows aggressive grading on the environment while keeping faces natural.
  3. Lifted blacks are the simplest path to "cinematic" — True black in video looks digital and harsh. Lifting the black point to 5-10% (so the darkest value is dark gray, not black) immediately creates the film-like latitude that audiences associate with cinema. One adjustment, dramatic effect.
  4. Consistency across a channel builds brand recognition — When every video has the same color grade, viewers recognize your content by its visual feel before reading the title. The grade becomes your visual signature — as distinctive as a logo.
  5. The same footage can tell different stories through grading — Warm grade: happy, inviting, trustworthy. Cool grade: professional, technological, precise. Desaturated: dramatic, serious, documentary. The color grade is not decoration — it is a narrative tool that shapes the viewer's emotional response to every frame.

Output Formats

Format Resolution Use Case
MP4 16:9 1080p / 4K YouTube / cinema / website
MP4 9:16 1080x1920 Social media
MP4 1:1 1080x1080 Instagram / LinkedIn
LUT (.cube) Reusable grade for editing software
PNG Before/after comparison strip

Related Skills

安全使用建议
This instruction-only skill appears coherent for an API-based video grading service, but the source is unknown and some metadata is inconsistent. Before installing or providing credentials: 1) Confirm the service endpoint(s) it calls and read the privacy/retention policy — your videos will likely be uploaded to an external server. 2) Ask the publisher for a homepage or source code so you can inspect behavior. 3) Verify why NEMO_TOKEN is needed and whether it can be scoped/limited; do not reuse a token that grants broad access to other accounts. 4) Confirm the skill only reads ~/.config/nemovideo/ (its own config) and nothing else in your home directory. 5) Test first with non-sensitive sample footage. 6) If you remove the skill, rotate/revoke the NEMO_TOKEN. If the publisher cannot provide clear endpoints, privacy terms, or source, treat the skill as higher risk and avoid uploading confidential videos.
功能分析
Type: OpenClaw Skill Name: ai-video-color-grading Version: 1.0.0 The skill bundle provides documentation and API usage examples for an AI-driven video color grading service. It correctly identifies its requirements for an API token (NEMO_TOKEN) and a configuration directory (~/.config/nemovideo/), and the provided curl example points to a plausible service domain (mega-api-prod.nemovideo.ai) without any signs of malicious intent or prompt injection.
能力评估
Purpose & Capability
The skill claims to call a NemoVideo service for color grading; the declared primary credential NEMO_TOKEN is consistent with an API-based service. However the registry metadata lists no required env vars while also naming a primaryEnv (NEMO_TOKEN) — a small inconsistency. The declared config path (~/.config/nemovideo/) is plausible for storing the service config, but because the skill has no source/homepage, this raises a transparency concern.
Instruction Scope
The SKILL.md (instruction-only) describes uploading video and describing the desired look — that aligns with the stated purpose. Because the file is truncated in the provided bundle, I cannot fully confirm there are no instructions that read unrelated files or environment variables. Expect the skill to transmit user video data to an external service (NemoVideo); confirm the destination and data retention policy before use.
Install Mechanism
No install spec and no code files — lowest install risk. The skill will not write new binaries or run an installer as part of its package.
Credentials
Requesting one primary credential (NEMO_TOKEN) is proportionate for an external API service. The metadata's empty requires.env list vs. primaryEnv mismatch should be clarified. The declared config path could contain other secrets if users have previously stored credentials there — confirm the skill only reads its own service config and nothing else.
Persistence & Privilege
The skill is not always-enabled and uses normal autonomous-invocation settings. It does not request elevated persistence or modification of other skills.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install ai-video-color-grading
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /ai-video-color-grading 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of AI Video Color Grading (v1.0.0): - Provides professional AI-driven color correction and grading for any video. - Supports a wide range of enhancements: exposure, white balance, contrast, color temperature, saturation, highlight/shadow recovery, and skin tone preservation. - Enables natural language prompts and creative presets for cinematic, branded, or mood-specific looks. - Handles multi-camera color matching and batch grading for large video libraries. - Designed for YouTube creators, filmmakers, event videographers, and corporate content teams seeking consistent, high-quality color grading.
元数据
Slug ai-video-color-grading
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Ai Video Color Grading 是什么?

Professional color correction and color grading for any video with AI — transform flat raw footage into cinematic, branded, or mood-specific visuals. NemoVid... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 137 次。

如何安装 Ai Video Color Grading?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-video-color-grading」即可一键安装,无需额外配置。

Ai Video Color Grading 是免费的吗?

是的,Ai Video Color Grading 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Ai Video Color Grading 支持哪些平台?

Ai Video Color Grading 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Ai Video Color Grading?

由 peandrover adam(@peand-rover)开发并维护,当前版本 v1.0.0。

💬 留言讨论