All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
Breaking Changes:
- Simplified command names:
/lazy-dev:*→/lazy * - Plugin name changed to "lazy" for marketplace
- Restructured repository: moved from
LAZY_DEV/.claude/to.claude/at root
- Inline task management in US-story.md files (no separate task files)
- Debug report generation when
/lazy reviewfinds issues /lazy fixcommand to apply fixes from review reports- Smart defaults for test/review requirements based on project structure
- Commands now use
/lazyprefix instead of/lazy-dev: - Plugin name: "lazy-dev" → "lazy"
- Repository structure: consolidated to root
.claude/directory - Review workflow: generates structured debug reports on failure
- Command discoverability with simpler
/lazynamespace - Review error handling with machine-readable debug reports
- Automatic fix routing by issue severity
- Better marketplace compatibility
Breaking Changes:
- Renamed project from "LAZY" to "LAZY_DEV Framework"
- Updated GitHub repository to MacroMan5/claude-code-workflow-plugins
- Restructured for Claude Code plugin marketplace distribution
Commands (8 total):
/lazy create-feature- Generate US-story + tasks with PM agent/lazy task-exec- Execute task with TDD + quality pipeline/lazy story-review- Review complete story and create PR/lazy story-fix-review- Apply review fixes/lazy documentation- Generate/update documentation/lazy cleanup- Remove dead code/lazy memory-graph- Manual memory persistence/lazy memory-check- Verify MCP connectivity
Agents (10 total):
- project-manager - Creates user stories and tasks
- task-enhancer - Enriches task descriptions
- coder - Implements features with TDD
- tester - Writes comprehensive tests
- reviewer - Reviews code at task level
- reviewer-story - Reviews complete stories
- refactor - Improves code quality
- documentation - Generates/updates docs
- cleanup - Removes dead code
- research - Fetches documentation and examples
Skills (17 total):
- Planning: story-traceability, task-slicer, ac-expander, brainstorming
- Development: test-driven-development, diff-scope-minimizer, subagent-driven-development
- Quality: code-review-request, finishing-a-development-branch
- Documentation: writing-skills, example-skill
- Integration: gh-issue-sync, git-worktrees, memory-graph, output-style-selector, context-packer, dispatching-parallel-agents
Hooks (8 implementations across 4 types):
- UserPromptSubmit: user_prompt_submit.py (context enrichment + memory detection)
- PreToolUse: pre_tool_use.py (safety checks)
- PostToolUse: post_tool_use_format.py, memory_suggestions.py, memory_router.py, log_events.py
- SessionStart: session_start.py
- Stop: stop.py (quality gate logging)
Documentation:
- CREDITS.md - Attribution to PROJECT_INSPIRATION sources
- hooks.json - Hook registry for plugin system
- Comprehensive README.md with quick start
- WORKFLOW.md - Complete workflow guide
- MEMORY.md - MCP Memory integration guide
- SUB_AGENTS.md - Agent specifications
- CLAUDE.md - Framework guide for Claude Code
Plugin Infrastructure:
- .claude-plugin/plugin.json - Plugin manifest
- .claude-plugin/marketplace.json - Marketplace catalog
- hooks.json - Hook documentation
- Plugin name: "lazy-dev-tools" → "lazy-dev"
- Version: 0.1.0 → 2.0.0
- Repository: therouxe/lazy → MacroMan5/claude-code-workflow-plugins
- Documentation restructured for production release
- Agent format: Updated to Anthropic's official YAML+Markdown pattern (context-based, no variable substitution)
- Memory system: Clarified as semi-automatic (AI-assisted suggestions, not fully automatic)
- README.md: Added pros/cons, alignment with Anthropic best practices, troubleshooting
- Quality pipeline: Better cross-OS compatibility
- Hook system: Standardized JSON format
- Agent descriptions: Clearer action-oriented triggers
- Hook count documentation (4 types, 8 implementations)
- Agent delegation pattern (context-based, not variable substitution)
- Cross-platform path handling in scripts
Features:
- First public release
- Skills: memory-graph, writing-skills, TDD, diff-scope-minimizer, planning/brainstorming
- Hooks: UserPromptSubmit (context, style, memory auto-trigger), PreToolUse (safety), PostToolUse (formatter + memory suggestions), Stop (quality gate)
- Commands: memory-graph, memory-check, create-feature, task-exec, story-review, documentation
- MCP:
.mcp.jsontemplate for@modelcontextprotocol/server-memory - Docs: framework overview, memory setup/plan, skills catalog
- STT Prompt Enhancer: documented pipeline and behavior
This is a major production release with:
- Complete Plugin Infrastructure - Ready for Claude Code marketplace
- Comprehensive Documentation - 5 core docs + detailed guides
- Full Agent System - 10 specialized agents with clear delegation
- Quality Pipeline - Fail-fast enforcement (format→lint→type→test)
- MCP Memory Integration - Semi-automatic persistence across sessions
- Proper Attribution - Credits to all PROJECT_INSPIRATION sources
# Add marketplace
/plugin marketplace add MacroMan5/claude-code-workflow-plugins
# Install plugin
/plugin install lazy-dev@MacroMan5
# Set environment
export ENRICHMENT_MODEL=claude-3-5-haiku
# Verify
/lazy memory-checkIf you were using v0.1.0:
- Backup your
.claude/directory - Remove old installation
- Install v2.0.0 via plugin system
- Copy any custom configurations from backup
Repository: https://github.com/MacroMan5/claude-code-workflow-plugins License: MIT Author: MacroMan5 (Therouxe)