Skills for Real Engineers. Straight from my .agents directory.
wan-scail-replacement
Agent skill fromartokun/comfyui-mcp
SCAIL-2 in-video character replacement on WAN 2.1. WanSCAILToVideo + SCAIL2ColoredMask + SAM3, the reference-image framing→scale rule, and the tuning/compositing pitfalls
How to install wan-scail-replacement
npx skills add artokun/comfyui-mcp# or add only this skill to Claude Code: copy the folder
plugin/skills/wan-scail-replacement/
# into
~/.claude/skills/wan-scail-replacement/ (all projects)
.claude/skills/wan-scail-replacement/ (one project)
1# SCAIL-2 In-Video Character Replacement (WAN 2.1)23SCAIL-2 (zai-org, on WAN 2.1 14B) replaces the person in a driving video with a4character you supply as a reference image, end-to-end, with no pose maps, and5with multi-character support. It is the successor to WAN Animate / motion6transfer for the "swap the subject, keep the motion" job. The official ComfyUI7template is `video_wan21_scail2_character_replacement_int8`, built around8`WanSCAILToVideo` (+ `SCAIL2ColoredMask`) with a SAM3 mask driving where the9character goes.1011This skill documents the non-obvious behaviours that cost a full multi-minute12render to discover. It is not a from-scratch graph. Start from the official13template and apply the guidance below.1415## The one rule that costs a re-render: reference framing controls SCALE1617**In `replacement_mode: true`, the reference image's FRAMING controls the output18character's SIZE as well as its appearance.** The SAM3 mask controls where the19character is placed; the reference image controls how large.2021Measured on a 720x1280 driving clip where the subject occupied ~30% of frame22height (subject bbox 363 px):2324| Reference framing | Person bbox in output | vs driving subject |25|---|---|---|26| Full-bleed portrait (person ~93% of frame) | 621 px | **1.71x oversized** |27| Reframed (person ~34% of frame) | 364 px | **1.003x — correct** |2829After reframing, top/bottom registration matched the driving subject within 1 px30…
Excerpt shown under the repository's MIT license.
More skills in comfyui-mcp
Similar skill repositories
Public repository for Agent Skills
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Production-grade engineering skills for AI coding agents.
Agent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
