v1.2.0 - Default values and JSON-native resolution in templating

What's Changed

• Support default values in input schema and skip optional properties without defaults by @Copilot in #64
• Implement JSON-native resolution for boolean, array, and object properties via {!! ... !!} templating by @Copilot in #66

Full Changelog: v1.1.4...v1.2

v1.1.4 - Fixed issue with asyncio

Full Changelog: v1.1.3...v1.1.4

Cleanup

v1.1.3

cleanup

Debug

v1.1.2

debug messages

Validating mode

What's Changed

• Add validating mode to MCIClient for pure schema validation without side effects by @Copilot in #62

Full Changelog: v1.1.0...v1.1.1

v1.1 Toolsets, MCP and more!

What's Changed

• Add YAML schema file support (.yaml, .yml) by @Copilot in #38
• Add path validation security to restrict arbitrary file access in tool execution by @Copilot in #40
• Add disabled field and annotations object to Tool schema (breaking) by @Copilot in #43
• Update execution result format for MCP compatibility with structured content by @Copilot in #46
• Add CLI executor metadata: exit_code, stdout_bytes, stderr_bytes by @Copilot in #48
• Add tags field and tag-based filtering to Tool model by @Copilot in #49
• Add toolsets feature: reusable tool libraries with schema and adapter-level filtering by @Copilot in #51
• Add LiteMcpClient: lightweight MCP client integration using official mcp package by @Copilot in #53
• Add MCP server integration with auto-caching, filtering, and advanced templating by @Copilot in #55
• Add MCP servers to schema reference and reorganize documentation with concept guides by @Copilot in #57
• Capture MCP tool annotations and convert to tags + Support string toolset names by @Copilot in #59

Full Changelog: v1.0.0...v1.1.0

v1.0 - Model Context Interface

MCI

Initial release of the Python adapter for standardised, unified, platform-agnostic interface for managing model context. A lightweight, simple way to build, maintain and share AI toolsets.

What's Changed

• Add mci-example.json with comprehensive execution type examples by @Copilot in #2
• Add comprehensive implementation plan for MCI v1 Python adapter by @Copilot in #6
• Add automated setup script and GitHub Copilot configuration for environment setup by @Copilot in #4
• Implement core infrastructure and data models with Pydantic validation by @Copilot in #8
• Implement templating engine with basic and advanced features by @Copilot in #10
• Implement Stage 3: Base, File, and Text Executors with Config-Level and Advanced Templating Support by @Copilot in #15
• Implement HTTP Executor with authentication, retry logic, and metadata support by @Copilot in #16
• Implement CLI Executor (CLIExecutor) with comprehensive templating and platform support by @Copilot in #17
• Implement ExecutorFactory and End-to-End Execution Feature Tests (Stage 3.3) by @Copilot in #18
• Implement SchemaParser (Stage 4.1) with comprehensive validation and testing by @Copilot in #21
• Implement Tool Manager (Stage 4.2) by @Copilot in #22
• Implement Stage 5: MCIClient - Main API for MCI Python Adapter by @Copilot in #24
• Add comprehensive Quickstart Guide for MCI Python Adapter (Stage 7.2) by @Copilot in #30
• Add comprehensive API documentation for MCIAdapter (Stage 7.1) by @Copilot in #28
• Add comprehensive schema reference documentation (Stage 7.3) by @Copilot in #32
• Add comprehensive example files for MCI Python Adapter (Stage 7.4) by @Copilot in #34
• Update README.md with comprehensive project overview, features, documentation, examples, support, and credits by @Copilot in #36

New Contributors

• @Copilot made their first contribution in #2

Full Changelog: https://github.com/Model-Context-Interface/mci-py/commits/v1.0.0