Add research skills for docs and Evaluation Guide

[jdiehl]

Adds two new AI assistant skills to support research workflows across the documentation repository and the Mendix Evaluation Guide.

Changes

docs-research skill (.claude/skills/docs-research/SKILL.md, .github/prompts/research.prompt.md)

• New skill for locating and summarizing content across the 4,500+ Markdown files in the docs repository
• Includes a topic map covering all major documentation sections (Studio Pro, mobile, deployment, Marketplace, etc.)
• Defines a consistent workflow: identify topics → find source files → extract relevant passages → respond with citations
• Companion Copilot prompt for the same research workflow

evalguide-research skill (.claude/skills/evalguide-research/SKILL.md, .github/prompts/evalguide-research.prompt.md)

• New skill for fetching and summarizing content from the Mendix Evaluation Guide at mendix.com/evaluation-guide/
• Covers platform capabilities, architecture, security, deployment, governance, and strategic partnerships
• Uses WebFetch to retrieve live content and returns direct answers with source citations
• Companion Copilot prompt for the same workflow

CLAUDE.md

• Updated the role description to mention research as a supported use case
• Added a full skill directory under the instruction precedence section so the agent knows which skill to invoke for each task type

[MarkvanMents]

Need to look at this skill and compare with Maia, which is much more token-efficient because of the RAG.
Will ask Esteban and team to look at this from their experience to see what the benefits are.
Feel free to review, but please do not merge until the Maia team have had a look at this skill.