NGSS MCP Server

# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [1.1.3] - 2025-10-20 ### Added - GitHub Actions automated publishing workflow with OIDC authentication - Automatic npm publishing on version tag push - GitHub Release creation with changelog integration - Build and test verification before publishing ### Infrastructure - No more manual npm publish required - fully automated via GitHub Actions - Secure OIDC authentication (no NPM_TOKEN needed in secrets) - Provenance attestation for supply chain security ## [1.1.2] - 2025-10-20 ### Fixed - Documentation: Corrected tool count from 9 to 8 (Tool 4 was removed) - Documentation: Corrected DCI count from 14 to 35 in README - Added prominent middle school standards disclaimer at top of README ## [1.1.1] - 2025-10-20 ### Fixed - **Critical Data Quality Fix**: Corrected DCI assignments across 29 standards - Expanded DCI_VALUES enum from 14 to 35 complete middle school DCIs - Fixed Unicode smart quote in "Earth Materials and Systems" DCI name - All 55 standards now have correct DCI mappings (32 of 35 DCIs mapped to standards) - Tool 8 (search_by_disciplinary_core_idea) now supports all 35 NGSS middle school DCIs ### Technical - DCI coverage: 14 → 35 DCIs (100% NGSS middle school coverage) - Standards corrected: 29 of 55 standards updated with correct DCIs - Test count: 99 tests passing (87 → 99, added 12 Tool 8 tests) - Missing DCIs (not mapped to standards): PS3.C, PS4.B, ESS2.E ## [1.1.0] - 2025-10-20 ### Added - **Tool 5: search_by_practice** - Filter standards by Science and Engineering Practice (10 SEP values) - **Tool 6: search_by_crosscutting_concept** - Filter standards by Crosscutting Concept (8 CCC values) - **Tool 8: get_unit_suggestions** - Intelligent unit planning recommendations with binary compatibility scoring - Complete 3D framework filtering support across all SEP and CCC dimensions - Binary compatibility scoring algorithm for unit planning (domain +3, SEP +2, CCC +2, DCI +1) - Data validation tests enforcing ADR-001 single-object model (sep.name, ccc.name, dci.name) - Backward compatibility regression tests for Tools 1-4 - Tool regression smoke tests for Tools 5, 6, 8 - Comprehensive integration test suite: 87 tests total (79 existing + 8 new) ### Changed - Data model enforces single objects per ADR-001 (not breaking - backward compatible) - Improved test coverage: 79 tests → 87 tests (100% code coverage maintained) - Enhanced README with Tools 5, 6, 8 documentation, SEP/CCC enum values, and ADR-001 alignment note - Updated project structure documentation to reflect 8 tools ### Technical - ADR-001: Single-object 3D framework structure for consistent data model - Test count: 87 comprehensive tests (54 baseline + 25 Tools 5-6-8 + 8 integration) - 100% backward compatibility verified (all 54 v1.0.1 baseline tests pass) - Package size: Optimized for 8 tools with minimal overhead ## [1.0.1] - 2025-01-19 ### Fixed - Remove duplicate domain enum values (6→3 domains) in search_by_domain tool - Fix pagination support in search_by_domain handler using array.slice() - Add descriptive examples to get_3d_components tool description - Add descriptive examples to search_standards tool description - Remove stale test for deleted find_by_driving_question tool - Update README installation instructions to use npx command ### Changed - Add `total` field to search_by_domain response for pagination context - Package size: 44.7 KB (60 files) ## [1.0.0] - 2025-01-19 ### Added - Initial release of NGSS MCP Server - 4 tools for accessing NGSS Middle School standards: - `search_by_domain`: Find standards by science domain - `get_3d_components`: Extract 3D framework components (SEP/DCI/CCC) - `search_standards`: Full-text search across all content - `get_standard_by_code`: Retrieve specific standard by code - Token-efficient data access with 3 detail levels (minimal/summary/full) - Query validation and caching for performance - Fuzzy search with Levenshtein distance for error tolerance - Complete NGSS MS standards dataset (Physical Science, Life Science, Earth and Space Science) - MCP protocol integration via @modelcontextprotocol/sdk ### Technical - TypeScript implementation with type safety - Optimized JSON data storage (106.3 KB) - Comprehensive test coverage - Token metadata tracking for LLM context management [1.0.1]: https://github.com/Sallvainian/NGSS-MCP/compare/v1.0.0...v1.0.1 [1.0.0]: https://github.com/Sallvainian/NGSS-MCP/releases/tag/v1.0.0