Run visual co-creation workspaces for coding agents with Pneuma Skills
Use Pneuma Skills when Claude Code, Codex, or Kimi should build slides, documents, diagrams, videos, dashboards, or custom modes in real files while a human reviews and edits through a live visual workspace.
npx skills add agentskillexchange/skills --skill run-visual-co-creation-workspaces-for-coding-agents-with-pneuma-skills
bunx pneuma-skills. For source installs, clone the repository, run bun install, and start a mode with commands such as bunx pneuma-skills slide --workspace ./my-first-pneuma-slide or bunx pneuma-skills doc --backend claude-code.Use Pneuma Skills when an agent/operator needs a repeatable visual co-creation workspace around file-based agent work, not when they simply want to chat with a coding agent or install a generic skill pack. The concrete workflow is to start a Pneuma mode such as slide, doc, diagram, remotion, gridboard, webcraft, or mode-maker against a workspace, choose a supported backend such as Claude Code, Codex, or Kimi, let the agent work in files, and review or edit the result through the mode-specific live viewer.
What this skill actually does
Invoke this instead of using Claude Code, Codex, or Kimi directly when the output needs a domain surface: slide preview and export, diagram canvas, document preview, video composition, dashboard tiles, or a custom mode with reusable skill context. The scope boundary is the mode-based visual collaboration loop around agent-authored artifacts. It is not a generic desktop app listing, marketplace card, broad framework entry, or all-purpose agent IDE.