mcp-server-index

name	mcp-server-index
description	RAG-indexed registry of configured MCP servers, available tools, and dispatch rules for Godot development workflows.

MCP Server Index

This skill is the source of truth for selecting MCP servers and tools in this environment. Always consult this index before invoking MCP tools.

Active MCP Servers

gopeak (Godot MCP)

• Source: HaD0Yun/Gopeak-godot-mcp
• Purpose: Godot runtime + project + scene operations, LSP, DAP debugging, screenshots, asset search, animations, audio, navigation, testing.
• Config key: gopeak
• Profile: compact (33 core tools + 22 dynamic groups with 78 additional tools)
• Typical use:
  • Run/stop project for validation.
  • Read runtime/debug output.
  • Programmatic scene and node operations.
  • GDScript LSP completions, hover, and symbols.
  • DAP breakpoints, stepping, stack traces.
  • Screenshots and viewport capture.
  • Animation, audio, tilemap, navigation management.
• Preferred tasks: runtime testing, scene refactors, verification loops, debugging.
• Note: Supersedes the older godot-mcp (Coding-Solo). The npm package gopeak provides both gopeak and godot-mcp CLI binaries.

context7

• Purpose: fast documentation retrieval.
• Config key: context7
• Typical use:
  • Pull current docs for Godot/GDScript/FastMCP/APIs.
• Preferred tasks: API lookups, syntax confirmation, library guidance.

godot-asset-library

• Purpose: search and pull assets from official Godot Asset Library API.
• Config key: godot-asset-library
• Tools:
  • get_categories
  • search_assets
  • get_asset
  • download_asset_zip
• Preferred tasks: find existing addons/assets before building from scratch.

kscript-knowledge

• Purpose: local RAG-like knowledge base for patterns, rules, references.
• Config key: kscript-knowledge
• Tools:
  • add_entry
  • search_entries
  • get_entry
  • list_tags
  • stats
• Preferred tasks: capture and retrieve reusable Godot/GDScript patterns.

ragx-aipm

• Purpose: AI project management and planning utilities.
• Config key: ragx-aipm
• Preferred tasks: planning and task organization support.

Dispatch Rules

1. Runtime errors / behavior mismatch -> gopeak first.
2. API uncertainty / docs lookup -> context7.
3. Need plugin/addon quickly -> godot-asset-library before custom implementation.
4. Reusable fix/pattern discovered -> store in kscript-knowledge via add_entry.
5. Multi-step roadmap -> ragx-aipm.

Godot Workflow Policy

• Start with analysis + minimal wireframe.
• Use MCP calls intentionally and explain why each call is needed.
• Keep changes backward compatible.
• Validate with runtime checks when possible.
• Persist useful lessons into kscript-knowledge.