Unofficial local MCP server that lets AI assistants draw in Aseprite
View on GitHub ↗Let an AI assistant draw live in your running Aseprite via the Model Context Protocol.
Unofficial community project. Not affiliated with or endorsed by Anthropic or Igara Studio (Aseprite). Use at your own risk – see DISCLAIMER.md.
"create me a map inside a dark medieval castle with dark wooden floor include a tilemap 16x16 for it"
"now draw me a wizard for the castle interior and create a walk animation in all four directions"
The assistant can create sprites, set exact pixels, use Aseprite's tools (line, rectangle, ellipse, bucket, curve, freehand …), manage layers and animation frames, undo/redo, and look at its own result as an upscaled PNG to iterate on it. Every change lands in Aseprite's undo history.
Drawing is designed to be cheap in tokens: whole sprites are sent as a palette plus text rows, several steps fit into one call, and pixels can be read back in the same text format.
- Aseprite v1.3 or later (Windows, macOS or Linux; the Steam version works too) - Node.js 20+ (some Linux distributions ship older versions – use nodejs.org or a version manager) - An MCP client that runs local (stdio) servers, e.g. Claude Code, Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf, Gemini CLI or Codex CLI. Clients that only support remote (HTTP) servers, such as ChatGPT, cannot use it.
Download aseprite-mcp-bridge.aseprite-extension from the releases (or build it with npm run build), then in Aseprite: Edit › Preferences › Extensions › Add Extension.
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/sloplink--aseprite-mcp.html)
Claude Code / Codex / pi plugin that hands agent steps needing no text output to Jev (TypeSafe's judgment model) — measured p50 ~230 ms and ~$0.02 per 1,000 judgments, with typed escalation back to th
neuronto/agentic-resource-discoveryNeuronto Agentic Resource Discovery (ARD) Index. Federated search across every public ARD registry, plus a verified tool index read from each MCP server's own tools/list, hybrid lexical and semantic r
PyModel/jev-judge-mcpTyped judgment tools for MCP agents. TypeSafe's Jev model as verify, screen, find, classify, rerank, decide, compare, extract, review, gate, and score: the model judges, policy decides auto, review, o
SysikNagibator/AiPCGive the agent "this" and he gets a real PC.
ZeroDayEvil/ai-security-tool🛡️ Free open-source AI-powered security terminal & vulnerability scanner (CVE, SBOM). Supports SSH, SFTP, RDP, VNC, Serial, and 12+ autonomous AI agents (DeepSeek, OpenAI) for automated security workf
taurus42119-stack/godkiller-mcpYour coding agent gaslights you. GODKILLER doesn't. Disk gates > vibes. claim_done only after verify.
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.