mcp_profile 0.1.0

AI persona definitions with template rendering, context injection, profile selection, and appraisal scoring for the MCP ecosystem.

0.1.0 Initial Release

Added

Core Features

• Profile Definitions

  • Profile model with system prompt templates
  • Metadata support (tags, categories, version)
  • Context variable specifications
  • Versioning and history tracking

• Profile Runtime

  • ProfileRuntime as the main execution engine
  • Profile registration and discovery
  • Template rendering with context injection
  • Execution tracking and metrics

• Template Rendering

  • Mustache-style template syntax
  • Context variable injection
  • Conditional sections and loops
  • Expression language support

• Profile Selection

  • Context-based profile selection
  • Confidence scoring for selections
  • Alternative suggestions
  • Selection criteria customization

• Appraisal System

  • Profile quality scoring
  • Feedback generation
  • Threshold-based validation
  • Custom appraisal criteria

• Port-Based Architecture

  • ProfileStoragePort for profile persistence
  • ProfileRenderPort for template rendering
  • ProfileSelectionPort for profile selection
  • ProfileAppraisalPort for quality scoring
  • FactGraphPortL1 for fact graph integration

• In-Memory Implementations

  • Complete in-memory storage for testing
  • Default render and selection implementations

Data Models

• Profile - Profile definition with templates
• ProfileContext - Rendering context data
• ProfileExecutionResult - Rendering outcome
• ProfileSelectionResult - Selection outcome
• ProfileAppraisal - Quality assessment

---

Support and Contributing

• Report Issues
• Join Discussions
• Read Documentation
• Support Development