Maintainer of this skill?
by delexw
Turn ideas into motion graphics and animated videos with Claude or Codex. 用 Claude 或 Codex，将创意变成动态图形和动画视频。
# Add to your Claude Code skills
git clone https://github.com/delexw/motion-graphics-agentGuides for using ai agents skills like motion-graphics-agent.
See how motion-graphics-agent compares with popular alternatives.
motion-graphics-agent is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by delexw. Turn ideas into motion graphics and animated videos with Claude or Codex. 用 Claude 或 Codex，将创意变成动态图形和动画视频。. It has 51 GitHub stars.
motion-graphics-agent's catalog security scan is still queued. You can run an instant dependency and prompt-injection check now with the "Scan for vulnerabilities" button above.
Clone the repository with "git clone https://github.com/delexw/motion-graphics-agent" and add it to your Claude Code skills directory (see the Installation section above).
motion-graphics-agent is primarily written in TypeScript. It is open-source under delexw on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh motion-graphics-agent against similar tools.
No comments yet. Be the first to share your thoughts!
Available after the catalog security scan passes.
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
The deep catalog scan for this skill is still queued. Run an instant dependency check now instead.
Use Kiteva from external agents through its MCP server: search ideas and Motion Graphics, save an Assistant draft, and explicitly start a build.
Switch the interface instantly with EN / 中文; see interface language. Video language remains a separate project setting.
Dates below use New Zealand time.
import_asset. Use manage_idea_assets to list available assets, attach them to the draft or remove their entries from the material shelf. MCP and the interface share file limits and attachment rules. Imported files are stored in Git-ignored runtime uploads. See agent MCP.start_idea creates a project from a saved draft, Kiteva keeps its attachment IDs and associates the imported files with that project. Updates reject outdated draft revisions and files belonging to another project./mcp endpoint. Tools share typed results and request validation, protect concurrent draft edits, and default to Claude Opus 5.5, Multiplane, 2K, 60 fps, Explore and three preview directions. See agent MCP.18419 for the interface and 18417 for the backend, with shared validation and matching proxies, origin checks and verification scripts. Invalid or colliding overrides fail explicitly; the frontend never silently takes another port.v1.0.0 source release: publish the production workflow improvements below under the renamed motion-graphics-agent repository. See the changelog.
Production Desk and inline storyboard: open both viewers inside centered, nearly full-window dialogs. Desk provides read-only playback, seeking, zoom, clip inspection and real audio waveforms; its larger preview leaves tracks below. Saved production metadata is available during generation, before the agent returns its accepted response. See Production Desk.
Separate audio sources: new full builds retain music-only and individual SFX files with timed placements and gains in a shared audioTracks manifest. The agent still prepares the final music/SFX mix through the existing renderer handoff; Desk inspects the separate sources rather than treating the merged soundtrack as music-only. Application-owned narration mixing is unchanged, and legacy projects keep their fallback.
Playback synchronization: move the Desk cursor with CSS transforms using the media clock, without frame-by-frame React updates or repeated layout positioning. Opening Desk pauses background audio; shared video players prevent competing audible playback while leaving silent previews running.
Ming narration: add the local Ming 3.0 TTS 0.5B provider, its supported-language voice catalog and neutral voice controls. Retire both Qwen providers, retain the original narrator reference for Ming cloning, and normalize saved legacy speech selections. Narration guidance now asks for a connected story with natural chapter transitions. See Ming TTS.
Shared project workspace: centralize artifact, asset, diagnostic and runtime paths for build agents, correctors, capture, narration, exports and cache recovery. Render inputs stay under artifacts; logs and task context remain separate. Existing workspaces and recovery checkpoints use explicit layout compatibility. See project workspace.
Cache evidence and correction outcomes: replace the Haiku change-summary step with bounded factual evidence for JEV, preserving hash checks and concrete cache-restore reporting. Corrections return structured fixed/refused/unresolved outcomes, escalate non-repairs to the main model and retain reasons for narrowly scoped bypasses. Renderer execution uses native defaults. See cache recovery and correction outcomes.
Consistent project controls: recover from a broken brief without losing the option to remove its draft, copy style metadata, and name downloads from the normalized style and version. Storyboard actions use matching icons and lowercase labels; activity grouping and render progress keep their task hierarchy.
Credential boundaries and verification: share credential environment names across SDK configuration, child processes and provider tests; deny credential access in agent shell tools while retaining required SDK authentication. Codebase tests clear provider credentials and block real external API calls. A separate live 10-second project verified the model-created music/SFX mix, saved source clips and Desk waveforms.