Skip to content

Releases: LLMdotInfo/ldf

LDF v1.2.0

Choose a tag to compare

@JayFromEpic JayFromEpic released this 30 Dec 08:11
Immutable release. Only release title and notes can be modified.
e647733

What's New

Python 3.14 Support

  • Added Python 3.13 and 3.14 to CI test matrix and pyproject classifiers
  • Updated dev dependencies: pytest>=8.0.0, black>=24.0.0, ruff>=0.5.0, mypy>=1.10.0

Exception Handling Improvements

  • Improved exception handling across 9 files to catch specific exceptions with proper logging
  • Added TypeError, KeyError handling for malformed YAML configurations
  • Prevents silent failures while maintaining graceful degradation

Bug Fixes

  • Fixed guardrail matrix parser to handle **Reference:** line in templates
  • Switched __version__ to importlib.metadata for single-source versioning

Testing

  • Added comprehensive tests for exception fallback scenarios
  • 1318 tests passing, 91.78% coverage

LDF v1.1.2

Choose a tag to compare

@JayFromEpic JayFromEpic released this 30 Dec 00:51
Immutable release. Only release title and notes can be modified.

What's New in v1.1.2

Bug Fixes & Improvements

  • Fixed all mypy type errors - Added proper type annotations throughout the codebase
  • Fixed guardrail matrix path - Corrected path to ldf/_framework/guardrails/ in CI workflow
  • Added pytest-asyncio - For async test support
  • Added types-PyYAML - For proper YAML type checking

Features from v1.1.1/v1.1.2

  • Task grouping in ldf tasks - Tasks are now grouped by phase
  • Cross-project section validation - Validates #section references in cross-project specs
  • Workspace targeting improvements - Fail-fast on explicit --project/--workspace

Documentation & Infrastructure

  • Added GitHub issue templates
  • Added SECURITY.md
  • Improved CI workflows with workflow_dispatch triggers

Full Changelog: v1.1.0...v1.1.2

LDF v1.0.0

Choose a tag to compare

@JayFromEpic JayFromEpic released this 26 Dec 20:13

LDF v1.0.0 - Initial Public Release

LLM Development Framework - Spec-driven development for AI-assisted engineering.

Features

  • CLI Tools: ldf init, lint, audit, create-spec, coverage, convert, update, doctor
  • Guardrails System: 8 core guardrails + 4 domain presets (saas, fintech, healthcare, api-only)
  • Question-Packs: security, testing, api-design, data-model
  • Template Engine: Jinja2 macros for clarify-first, coverage-gate, task-guardrails
  • MCP Server Integration: spec_inspector, coverage_reporter for IDE tooling
  • Multi-Agent Audit: ChatGPT and Gemini integration for spec review
  • Security Hardening: Path traversal prevention, symlink validation

Installation

pip install ldf

Documentation

See the README for quick start guide and full documentation.

License

MIT