mcp-tool-selection

Choose between MCP semantic search and literal search for code exploration.

402|54|Updated Oct 8, 2025
One-click install
npx skills add https://github.com/Context-Engine-AI/Context-Engine --skill mcp-tool-selection-context-engine-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-tool-selection
Source: https://github.com/Context-Engine-AI/Context-Engine/tree/main/.skills/mcp-tool-selection
Command: npx skills add https://github.com/Context-Engine-AI/Context-Engine --skill mcp-tool-selection-context-engine-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Decision rules for when to use MCP Qdrant-Indexer semantic search vs grep/literal file tools. Use this skill when starting exploration, debugging, or answering "where/why" questions about code.

Core Features & Use Cases

  • Provides a clear decision framework that prioritizes semantic MCP tools for exploration, debugging, and architectural understanding.
  • Specifies when to use literal search/file-open for exact-match lookups and quick checks.
  • Includes STOP rules and a quick decision heuristic to streamline code-reasoning workflows.

Quick Start

Ask MCP to guide your exploration strategy before performing any code searches.

Frequently Asked Questions about mcp-tool-selection

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

FAQPage Schema
What's the best way to search for code definitions across multiple repositories?

To search code definitions across multiple repositories, use MCP Qdrant-Indexer semantic search for understanding architectural context, and use literal search tools like grep for exact string matches. This decision framework ensures safe, efficient exploration workflows without unnecessary file reads.

How do I decide between semantic search and grep for debugging code?

When debugging code, prioritize MCP Qdrant-Indexer semantic search to locate definitions and trace code paths for conceptual understanding, while using literal search tools for exact-match lookups and quick checks. Clear decision rules and STOP heuristics streamline the code-reasoning process.

When do I need semantic code search instead of literal text matching?

You need semantic code search when answering where or why questions across repositories that require architectural understanding, whereas literal text matching is suited for exact-match lookups and quick checks during code exploration.

How can I prevent unnecessary file reads when tracing code paths?

To prevent unnecessary file reads while tracing code paths, apply the STOP rules and quick decision heuristic provided by the tool-selection framework to enforce a safe, efficient exploration workflow using MCP semantic search and literal search appropriately.

Does the MCP Qdrant-Indexer semantic search work for single repository code exploration?

Yes, the MCP Qdrant-Indexer semantic search applies to code exploration and debugging across one or more repositories, prioritizing semantic understanding for locating definitions and tracing code paths while reserving literal search for exact matches.