skills.rest
Mathews-Tom
Mathews-TomCOMMUNITY
Explore all skills by this author →
𝕏@tom_mathewsGitHub

mcp-to-skill

Convert MCP server tool schemas into validated SKILL.md packages with token-savings estimates.

310|45|Updated Feb 22, 2026
Search Tags:#automation#cli#mcp#conversion#token-optimization#rest-api#skill-generation
One-click install
npx skills add https://github.com/Mathews-Tom/praxis-skills --skill mcp-to-skill-mathews-tom
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mcp-to-skill
Source: https://github.com/Mathews-Tom/praxis-skills/tree/main/skills/mcp-to-skill
Command: npx skills add https://github.com/Mathews-Tom/praxis-skills --skill mcp-to-skill-mathews-tom

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

MCP-to-Skill addresses excessive context token usage caused by MCP server tool schemas being injected into every turn. It helps teams reclaim context window capacity by converting stateless, high-token-cost MCP tools into on-demand skills that load only when needed.

Core Features & Use Cases

  • Tool Discovery & Inventory: Extracts tool definitions from active sessions, config files, repo source, package registries, or user-provided schemas.
  • Classification & Strategy: Classifies tools (REST_API, CLI_WRAPPER, LOGIC_PATTERN, FILE_OP, STATEFUL, COMPOSITE) and recommends convert/keep/hybrid approaches with usage-frequency guidance.
  • Skill Generation & Validation: Produces a complete SKILL.md package, quick reference mappings, prerequisites, replacement commands or methodologies, token-savings estimates, and test scenarios.
  • Environment-aware Output: Includes environment guidance and scripts for token estimation and validation tailored to Claude.ai, Claude Code, and IDE agents.
  • Use Case Example: Migrate a Slack or GitHub MCP that exposes many stateless tools to a smaller set of on-demand skills to save thousands of tokens per turn.

Quick Start

Use the mcp-to-skill skill to analyze an MCP server's tool definitions, classify each tool by replacement strategy, and generate a validated SKILL.md package with estimated token savings.

Frequently Asked Questions about mcp-to-skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce context window token usage from MCP server tool schemas?▼

MCP servers expose tool schemas that get injected into every conversation turn, consuming excessive context window tokens. This bloats the context window and reduces available capacity for actual task content, especially when stateless REST API or CLI wrapper tools carry large schema definitions.

How do I convert MCP server tools into on-demand skills?▼

The process involves extracting tool definitions from active sessions or config files, classifying each tool as REST_API, CLI_WRAPPER, LOGIC_PATTERN, FILE_OP, or COMPOSITE, then generating a SKILL.md package with replacement commands, prerequisites, and token-savings estimates.

Can I use this for MCP servers that expose REST APIs and CLI wrappers?▼

Yes, this supports MCP servers exposing REST APIs, CLI wrappers, logic patterns, file operations, and composite workflows. It classifies tools by type and recommends convert, keep, or hybrid strategies with usage-frequency guidance tailored to each category.

What's the best way to migrate a Slack or GitHub MCP server to reduce tokens?▼

Migrating a Slack or GitHub MCP involves analyzing tool definitions, classifying stateless tools by replacement strategy, and generating a smaller set of on-demand skills with a SKILL.md package. This can save thousands of tokens per turn by loading schemas only when needed.

Does this work with Claude Code and IDE agents for token estimation?▼

Yes, environment-aware output includes scripts and guidance for token estimation and validation tailored to Claude.ai, Claude Code, and IDE agents. The generated SKILL.md package contains test scenarios and token-savings estimates validated for each specific environment.

When should I not convert an MCP tool into an on-demand skill?▼

Stateful tools and frequently used tools should be kept as MCP tools rather than converted. The classification strategy identifies STATEFUL tools as keep candidates, while stateless, high-token-cost tools with lower usage frequency are recommended for conversion to on-demand skills.

Related Skills

Explore More Software Engineering Skills→
cyanheadscyanheads
148

add-tool

Generate TypeScript MCP tool definitions and register them in the project.

Community
Intermediate
galeepgaleep

caveman-stats

Analyze Claude Code session logs for actual token usage and savings.

Community
Basic
obsfxobsfx

trekker

Track tasks persistently across sessions using Trekker as the source of truth.

Community
Intermediate
AradotsoAradotso
7

mcp-code-execution-mode

Proxy MCP tool calls through a sandboxed run_python bridge.

Official
Advanced