Add nonprofit-rbm-logic-model skill to .experimental

[vassiliylakhonin]

Summary

Add a new skill: nonprofit-rbm-logic-model under skills/.experimental/.

The skill generates Results-Based Management (RBM) logic models for nonprofit programs, with:

• 5-level results chain (Inputs -> Activities -> Outputs -> Outcomes -> Impact)
• SMART outcome indicators
• SDG alignment
• Monitoring/data collection recommendations

Files Added

• skills/.experimental/nonprofit-rbm-logic-model/LICENSE.txt
• skills/.experimental/nonprofit-rbm-logic-model/SKILL.md
• skills/.experimental/nonprofit-rbm-logic-model/agents/openai.yaml
• skills/.experimental/nonprofit-rbm-logic-model/references/rbm-framework.md

Validation

• Skill frontmatter includes name and description
• Skill name uses hyphen-case
• Skill validated locally with quick_validate.py
• agents/openai.yaml includes UI metadata and default prompt
• References are separated for progressive disclosure

Scope

Optimized for nonprofit/NGO program design and donor-ready M&E framing (USAID/UN/EU style reporting).

Example prompt

Use $nonprofit-rbm-logic-model to create a logic model for a girls education program in rural India.

[vassiliylakhonin]

Updated the skill based on testing feedback.

What changed:

• Refined Output Format in SKILL.md to start with Theory of Change.
• New required order is now:
  1. Theory of Change
  2. Executive Summary
  3. Logic Model Table
  4. Outcome Indicators
  5. SDG Alignment
  6. Data Collection Plan
  7. Assumptions and Risks

This makes the output align better with UN-style RBM framing before presenting the detailed results chain.