tool-usage-discipline

Prioritize high-context MCP tools and loaded skills for efficient tool selection.

1|Updated Mar 5, 2026

Search Tags:#workflow orchestration #knowledge base #code search #efficiency #token optimization #tool selection

Ready to run

npx skills add https://github.com/baphled/dotopencode --skill tool-usage-discipline

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: tool-usage-discipline
Source: https://github.com/baphled/dotopencode/tree/main/skills/tool-usage-discipline
Command: npx skills add https://github.com/baphled/dotopencode --skill tool-usage-discipline

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents inefficient manual workarounds and redundant tool usage by guiding the selection of the most appropriate and efficient tools for a given task.

Core Features & Use Cases

Prioritizes MCP Tools: Encourages the use of specialized tools like LSP and grep over generic shell commands.
Leverages Skills: Promotes consulting loaded skills for domain knowledge before resorting to external searches.
Avoids Redundancy: Prevents re-fetching information already present in the context.
Use Case: When asked to find a specific function in a large codebase, this skill would guide the AI to use lsp_goto_definition or grep instead of manually browsing files.

Quick Start

Use the tool-usage-discipline skill to determine the best approach for searching code.

Frequently Asked Questions about tool-usage-discipline

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

FAQPage Schema

How do I optimize token usage when searching code in a large codebase?▼

To optimize token usage during code search, prioritize high-context MCP tools like `grep` or `lsp_goto_definition` over manually browsing files to prevent context bloat. This approach ensures efficient session management by fetching precise definitions instead of loading entire files.

What is the best way to avoid reinventing the wheel with manual exploration?▼

The best way to avoid reinventing the wheel is to consult loaded skills for domain knowledge before resorting to external searches. This prevents redundant tool usage and leverages existing context to guide intelligent tool invocation for the task.

How do I prevent redundant tool invocation during workflow orchestration?▼

Prevent redundant tool invocation by avoiding the re-fetching of information already present in the context. Guide your tool selection to use specialized tools once, satisfying the need for optimized token usage without repeating external searches.

Why does using generic shell commands for code search increase context bloat?▼

Using generic shell commands increases context bloat because they often return unstructured or excessive output, whereas specialized MCP tools like LSP provide targeted results. Prioritizing these high-context tools reduces unnecessary token consumption.

When do I need to prioritize MCP tools over loaded skills for a task?▼

Prioritize MCP tools over loaded skills when domain knowledge is insufficient for the specific task, such as finding a function in a large codebase. Use LSP or grep to execute the search efficiently without manual exploration.