color-grading-finishing
Provider-independent color grading and finishing direction for generated video, ads, trailers, product films, social clips, explainers, and mixed-source edits. Use when planning, directing, reviewing, or QAing color correction, shot matching, exposure, contrast, saturation, skin/product color protection, look development, LUT/reference use, SDR/HDR color management, delivery constraints, generated-media artifacts, editor/compositor handoff, and finishing QC.
npx skills add calesthio/generative-media-skills --skill color-grading-finishing --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Guides the color and finishing stage of a video project, applicable to AI-generated, camera-originated, motion graphics, screen capture, avatar footage, product renders, or mixed-source edits. It emphasizes treating color as both a technical pipeline and storytelling surface, focusing on creating a cohesive finish that preserves required colors, supports intended emotion, survives compression, and gives downstream editors unambiguous instructions.
How it works
- Start by separating the job into three layers: Color correction (normalize and balance exposure, white balance, black/white points, contrast, saturation), Color grading (create the intended look after balance), Finishing (prepare final master for review/delivery with checks on legal/video levels, gamut, SDR/HDR target, accessibility, loudness, safe areas, artifact repair, and handoff materials).
- State evidence quality: separate documented facts (standards, platform specs, delivery targets, official tool behavior) from empirical observations and production heuristics.
- Decide the color-management path before grading: identify sources, target master, workflow, transforms, and review environment; preserve documented facts (e.g., ITU-R BT.709 for SDR, ITU-R BT.2100 for HDR, ACES IO transforms, OpenColorIO).
- Make a grading brief before touching controls: specify target deliverables, reference state, protections, scene families, look intent, and review scope.
- Follow a correction workflow in order: conform/tag, normalize sources, primary balance, shot match, secondary fixes, look pass, artifact pass, delivery/QC.
- For generated/mixed edits, emphasize perceptual continuity across cuts, with priorities on cut continuity, subject priority, scene logic, graphic stability, and motion consistency; note research context on spatiotemporal consistency and implications for frame-level repair if needed.
- Address skin, product, brand, and claim-sensitive colors to preserve identity; protect brand surfaces with isolated corrections when needed.
- Develop looks with concrete decisions on contrast architecture, palette, density, texture, and attention control, using look references and technical references as constraints.
- Clarify LUT roles (technical transform, creative look, viewing LUT) and cautions about stacking LUTs, inappropriate conversions, and providing appropriate review assets.
- Manage HDR/SDR and multiple masters with guidance on target delivery and platform considerations; provide caveats when HDR review displays or requirements aren’t available.
- Check platform and delivery constraints close to delivery, citing sources like YouTube SDR practices, TikTok ad specs, and general broadcast requirements; production heuristic suggests exporting a clean master plus platform-specific derivatives and verifying derivatives after compression.
- Ensure accessibility and viewer safety by preserving captions, UI labels, and color contrast guidance per WCAG references.
When to use it
Not explicitly listing triggers beyond general workflow; use when planning, directing, reviewing, or QAing color correction, shot matching, exposure, contrast, saturation, skin/product color protection, look development, LUT/reference use, SDR/HDR color management, delivery constraints, generated-media artifacts, editor/compositor handoff, and finishing QC.
What it can touch
The skill references several standards and workflows but does not specify executable tools beyond general workflow guidance; declared tools are: claude-code, codex, copilot, cursor. It mentions standards and sources (e.g., ITU-R BT.709, ITU-R BT.2100, ACES, OpenColorIO) but does not specify direct commands to run within these tools.
Caveats
- Contains references to standards and resources (e.g., ITU-R BT.709, ITU-R BT.2100, ACES, OpenColorIO) with verified dates.
- Production heuristic and platform notes are provided as guidance but should be verified against current specs before applying to a project.
- No outcomes promised; advice remains instructional and constraint-based, not prescriptive guarantees.
# Color grading and finishing direction Use this skill to guide the color and finishing stage of a video project, whether the material is fully AI-generated, camera-originated, motion graphics, screen capture, avatar footage, product renders, or a mixed-source edit. Treat color as both a technical pipeline and a storytelling surface. A good finish makes shots belong together, preserves required co
What does the color-grading-finishing skill do?
Provider-independent color grading and finishing direction for generated video, ads, trailers, product films, social clips, explainers, and mixed-source edits. Use when planning, directing, reviewing, or QAing color correction, shot matching, exposure, contrast, saturation, skin/product color protection, look development, LUT/reference use, SDR/HDR color management, delivery constraints, generated-media artifacts, editor/compositor handoff, and finishing QC.
How do I install it?
Run `npx skills add calesthio/generative-media-skills --skill color-grading-finishing --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From calesthio/generative-media-skills, a repository with 112 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.