character-replacement — Claude Code skill: install and use
character-replacement is a Claude Code skill that plans video character replacement, writes Ref2VA prompts, and configures ComfyUI workflows. It runs around segmented generation and review.
Apache-2.0 ★ 413 updated 2026-11-17
npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a claude-codeWhat character-replacement does
character-replacement plans character replacement from source images, video, and an explicit request. A vision-capable model inspects the footage, marks shot boundaries and each selected person's appearances, interactions, occlusion, and exit, then produces a segment plan with a prompt for every segment.
A single-person edit uses the H3-04 Character Replacement workflow. Multi-segment or multi-person work gets a task-specific copy of H3-04 Agentic Continuation. When generation is requested, the executable graph goes through ComfyUI, existing vLLM-Omni nodes, and the vLLM-Omni video API.
The setup needs a ComfyUI URL, available nodes, effective Omni deployment settings, and a task output directory. The source is copied to constant 24 FPS before inspection unless another rate is requested; the original file and playback timing stay intact. Prompt formatting depends on the MiniMax H3 skill, readable at https://github.com/MiniMax-AI/MiniMax-H3/tree/main/skills/h3-prompt-writing; local installation is optional.
What it helps with
- plan character replacement across video segments
- write per-segment Ref2VA prompts
- configure an H3-04 workflow for one person
- prepare multi-segment processing through ComfyUI
- review segment previews and the assembled video
Example prompts for the agent
- Plan character replacement in
source.mp4and write prompts for every segment. - Configure the H3-04 workflow for a single-person edit and prepare the ComfyUI settings.
- Inspect the faulty segment and revise its prompt without changing the boundary unless the footage shows it is wrong.
How to install character-replacement
1Skills CLI (Vercel)
npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a claude-code2Manually: put the files in place
~/.claude/skills/character-replacement/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/character-replacement/
1Skills CLI (Vercel)
npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a codex2Manually: put the files in place
.agents/skills/character-replacement/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a cursor2Manually: put the files in place
.cursor/skills/character-replacement/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ThinkFlowLab/OpenDirector packages/agent/skills/character-replacement2Manually: put the files in place
.github/skills/character-replacement/SKILL.md1Skills CLI (Vercel)
npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a gemini-cli2Platform CLI
gemini skills install https://github.com/ThinkFlowLab/OpenDirector --path packages/agent/skills/character-replacementManually: put the files in place
.devin/skills/character-replacement/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about character-replacement
How do I install character-replacement in Claude Code?
Run npx skills add ThinkFlowLab/OpenDirector --skill character-replacement -a claude-code, then use ~/.claude/skills/character-replacement/SKILL.md.
What does character-replacement do?
It inspects source video, builds a segment plan, writes Ref2VA prompts, and configures reusable H3-04 workflows for character replacement generation and review.
What does character-replacement need to run?
It needs source images or video, an explicit prompt, a ComfyUI URL, available nodes, Omni deployment settings, and an output directory. Prompt formatting uses the MiniMax H3 skill at https://github.com/MiniMax-AI/MiniMax-H3/tree/main/skills/h3-prompt-writing; generation passes through vLLM-Omni nodes and the vLLM-Omni video API.
Source text
SKILL.md · Apache-2.0
--- name: character-replacement description: Plan MiniMax-H3 character replacement, write per-segment Ref2VA prompts and configure reusable H3-04 ComfyUI workflows for generation and review. --- # Character Replacement Make character replacement easy to use: the user supplies source media and a replacement request; the agent derives the segment plan, prompts and workflow parameters. Keep the input boundary as original images, source video and explicit prompts. Work only on the requested deliverable; planning or prompt edits do not start generation, restart services or introduce preprocessing. For a simple one-person edit without segmented continuation, reuse the [H3-04 Character Replacement workflow](assets/character-replacement.json). For planned multi-segment work, use the Agentic Continuation template linked below; its example-specific settings must be recalculated for new footage. 1. Create a 24-FPS constant-frame-rate working copy of the source video before visual inspection, segmentation and frame-index calculations, unless the user explicitly requests another FPS. Preserve the original file and playback timing. Read [video-segment.md](video-segment.md) first. Use a vision-capable model to inspect the source video, establish shot boundaries and track when each selected person appears, interacts, becomes occluded and leaves. Produce the segment plan before writing new segment prompts. 2. Read [prompt-writing.md](prompt-writing.md), then write each segment's prompt from its inspected action and explicit source-person-to-picture mappings. 3. For a single-person, single-segment edit, configure the simple workflow linked above. For segmented or multi-person work, configure a task-specific copy of the [H3-04 Agentic Continuation workflow](assets/character-replacement-continuation.json), reusing Global Control, existing nodes and one shared segment-subgraph definition. Follow [workflow.md](workflow.md) for configuration, execution and review. 4. When generation is requested, submit the executable graph through ComfyUI, monitor results, inspect segment previews and deliver the assembled video with remaining quality defects. Prompt-format dependency: read the official [MiniMax H3 prompt-writing skill](https://github.com/MiniMax-AI/MiniMax-H3/tree/main/skills/h3-prompt-writing). It can be read directly from the repository; local installation is optional. Before execution, establish the ComfyUI URL, installed node availability, effective Omni deployment settings and task output directory as described in [workflow.md](workflow.md). Resolve repository-relative references from this skill directory after resolving any global skill symlink, not from the agent's current working directory. Execution boundary: agent → ComfyUI execution API → existing vLLM-Omni nodes → vLLM-Omni video API. The nodes own model requests; the agent does not bypass ComfyUI with a separate generation client. For a prompt-only correction to an existing segment, inspect its relevant footage and existing plan; revise the plan only if the observed defect exposes an incorrect boundary or mapping. Do not invent a new segmentation for every wording edit. Preserve the user's model, resolution, FPS and fixed seed unless asked to change them. Report written plans and prompts separately from generation and visual acceptance. Keep Markdown paragraphs on single source lines and usage notes brief.
© ThinkFlowLab · Apache-2.0 · ThinkFlowLab/OpenDirector