thierryc/litsquare-stage-integrations/litsquare-stage-render-video
Claude Code and agent skill by thierryc/litsquare-stage-integrations
- community
Checking that the SKILL.md file is still in the repository.
Render and natively encode LitSquare Stage projects through the mandatory macOS app. Use for app preflight, still capture, explicitly requested PNG sequences, final H.264 MP4/MOV encoding, motion blur render settings, render queue controls, diagnostics, inline render progress, artifact QA, and render troubleshooting.
Install
npx vskill@latest install thierryc/litsquare-stage-integrations/litsquare-stage-render-videoSKILL.md
LitSquare Stage Render Video
Use for every still, sequence, video, queue, progress, and diagnostics operation.
All fallback scripts use scripts/stage-mcp-client.mjs, initialize once, and call standard JSON-RPC MCP. Never construct or send the deprecated raw {tool,payload} envelope.
Mandatory Preflight
Use MCP-first preflight before every render:
- If the canonical
litsquare_stage_render_progresstool is available, call it without ajobID. - Treat a valid structured progress response, including
idle, as authoritative proof that the LitSquare Stage service and plugin MCP transport are ready. - Confirm that
litsquare_stage_start_video_render,litsquare_stage_start_sequence_render, andlitsquare_stage_set_render_window_stateare available before the corresponding workflow needs them. - Do not run the shell preflight when the canonical MCP tools are available.
Only when direct MCP tools are unavailable, run:
For client registration or transport problems, read references/mcp-client-integration.md before changing MCP configuration.
The fallback script requires localhost network access. If it reports service_unreachable for an installed or running app from a sandboxed command:
- Do not tell the user to restart the app based only on that shell failure.
- Prefer the direct MCP tool result.
- If direct MCP tools are absent, request permission for a read-only localhost probe or report that the client has not loaded the plugin MCP server.
Passing readiness requires macOS plus a healthy native litsquare-stage-macos service and the canonical render tools. Treat debug_app_unregistered as a non-blocking developer warning when the verified native service is healthy.
Do not use Chromium, Playwright, browser automation, or a remote service as a fallback.
Native Final Video Encoding
For every requested video deliverable, use litsquare_stage_start_video_render or the app-backed CLI with --kind video. LitSquare Stage owns frame rendering, motion-blur sampling, audio rendering, H.264 encoding, MP4/MOV container creation, muxing, and final metadata.
- Never create a final video by rendering a PNG sequence and passing it to FFmpeg or another encoder.
- Never use
ffmpeg, a browser recorder, an AVFoundation helper script, or another external encoder for transcoding, scaling, padding, frame-rate conversion, audio muxing, metadata, or container creation. - Use
litsquare_stage_start_sequence_renderonly when the user explicitly requests an image sequence as the deliverable. - For LinkedIn, Instagram, Bluesky, or any responsive pack, submit one native LitSquare Stage video render job per target size and output path. Do not render one master and transcode the variants externally.
- Set dimensions, fps, motion blur, codec/container, audio, and metadata in
stage.config.jsonor the LitSquare Stage render request. Do not repair these properties in post-processing. - If the app cannot produce the requested codec or container, stop and report that limitation. Do not silently substitute an external encoder.
- Verify completion from the app terminal progress state and returned video artifact. Use app-reported metadata or read-only file inspection for QA; do not mutate the artifact after the app returns it.
Render Workflow
- Inspect
stage.config.json. - If audio is requested, read
references/audio-sync.mdand confirmrender.audioEnabledistruein every target configuration. - Build the project using its package manager.
- Run
validate-stage-project.mjs. - Capture at least one still before long video or blurred renders.
- Preserve the user's window visibility and focus during rendering:
- The native renderer uses one compact layout for stills, sequences, and video.
- Show or bring it forward only when the user requests visible progress or inspection; use
compactMode: truefor compatibility. - Set
floating: trueonly when the user asks to keep the renderer on top. - Background rendering must finish without requiring foreground activation.
- For video output, call
litsquare_stage_start_video_renderdirectly. Do not call the sequence tool as an intermediate video-encoding step. - Use
litsquare_stage_start_sequence_renderonly for an explicitly requested image-sequence deliverable. - For long visible renders, poll
litsquare_stage_render_progresswith the returnedjobIDuntil that job is terminal. Treat its status and final events as authoritative; do not carry diagnostics from an older job into a retry. - Use
scripts/render-stage-project.mjs --kind videoonly when canonical MCP tools are unavailable or CLI automation is required. - Verify artifact existence, non-zero size, dimensions, frame count or duration, fps, visible content, native-app provenance, and requested audio tracks.
- Hand off to
litsquare-stage-quality-review.
Canonical quick-start request:
Motion Blur
Read references/motion-blur.md before changing blur settings. Draft renders should use no blur or low samples; final renders default to sampleCount: 8, shutterAngle: 180.
Output
Report project path, render settings, artifact paths, preflight status, progress mechanism, diagnostics warnings, and quality-review handoff.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/litsquare-stage-animation/skills/litsquare-stage-render-video/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/thierryc/litsquare-stage-integrations/litsquare-stage-render-video)