toolkit-awareness

Ground tool queries in the project's installed toolkit and documented commands.

6|1|Updated Jan 2, 2026
One-click install
npx skills add https://github.com/1cFE/agentic-mbse --skill toolkit-awareness
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: toolkit-awareness
Source: https://github.com/1cFE/agentic-mbse/tree/main/claude/skills/toolkit-awareness
Command: npx skills add https://github.com/1cFE/agentic-mbse --skill toolkit-awareness

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ensures Claude uses the actual installed toolchain when answering questions about validation, workflow steps, and available commands, preventing guesswork and inconsistencies.

Core Features & Use Cases

  • Grounded tooling: Routes all tool-related guidance to the project's README.md, CLAUDE.md, and the official toolchain.
  • Pre-answer checks: Verifies references to the four-directory architecture (knowledge, modeling_project, work, data) before responding.
  • No invention policy: Refuses to invent commands or steps not documented in the repository and cites the authoritative sources.

Quick Start

Consult the project README.md and CLAUDE.md before answering tooling questions to ensure commands and workflows come from the installed toolkit.

Frequently Asked Questions about toolkit-awareness

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

FAQPage Schema
How do I validate Claude CLI commands against the installed project toolkit?

To validate Claude CLI commands against the installed project toolkit, the skill consults README.md and CLAUDE.md to reference the actual toolchain, verifying commands and workflows before responding to prevent guesswork and ensure accuracy.

What is the best way to ground workflow steps in a project's actual toolchain?

Grounding workflow steps requires checking the project's README.md, CLAUDE.md, and the four-directory architecture (knowledge, modeling_project, work, data) to route all tool-related guidance to the official installed toolchain instead of guessing.

Can I use this approach to find available commands without reading the README?

Yes, you can use this approach to find available commands because it automatically queries README.md and CLAUDE.md to extract the official toolchain references, ensuring the commands provided are documented in the repository rather than invented.

Why does Claude invent CLI commands instead of using my installed toolkit?

Claude may invent CLI commands without a grounding mechanism to check the installed toolkit. This skill enforces a no-invention policy that refuses to generate commands or workflow steps not documented in README.md or CLAUDE.md.

Does this work with a four-directory project architecture?

Yes, it explicitly works with a four-directory project architecture containing knowledge, modeling_project, work, and data folders, requiring verification of these directories before responding to any tooling or validation questions.