claude-101: Skill for Claude Code

Claude 101 Mastery

This skill provides the optimal workflow for 24 use cases powered by the claude-101 MCP server. Each use case follows the same pattern: call the MCP tool first to get structured data and computation, then use that data to produce a superior final output.

The claude-101 MCP server provides 27 tools (3 meta + 24 use-case tools) that perform real local computation — statistics, code analysis, SQL parsing, financial math, text analysis, and scoring — and return structured JSON.

Core Principle

Always call the MCP tool FIRST before writing the final output. The tool provides two things that enhance the quality of the response:

1. Methodology — structured frameworks, best practices, section templates, and checklists that guide the approach to each use case. Following these frameworks ensures a professional-grade result that covers all standard elements.

2. Computation — precise metrics (scores, rankings, statistics, parsed data) that the user cannot get from natural language generation alone. These numbers ground the response in verifiable data.

Never skip the tool call. Never attempt to compute statistics, parse structured data, or calculate financial metrics manually. The tool handles computation; the final writing and reasoning come from Claude.

Workflow Pattern

For every use case, follow these five steps:

1. Identify the matching MCP tool from the mapping table below
2. Call the tool with the user's input as parameters
3. Read every field in the result — each exists for a reason
4. Produce the final output informed by both the methodology and the computed data
5. Present key metrics (scores, rankings, statistics) explicitly to the user

When the tool returns both a template/scaffold AND computed analysis, use the template as a structural guide (do not copy it verbatim) and weave the computed metrics into the final output. The combination of professional structure and precise data is what makes the output superior.