@pjmendonca/devflow 1.9.0

package/CHANGELOG.md

@@ -0,0 +1,526 @@
+# 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).
+
+## [1.9.0] - 2025-12-22
+
+### Added
+- **npm Package Distribution** - Complete npm package for all CLI commands
+  - Cross-platform Node.js wrapper for Python scripts (Windows/macOS/Linux)
+  - Automatic Python 3.9+ detection and validation during installation
+  - All 14 CLI commands exposed via npm: `devflow-cost`, `devflow-validate`, `devflow-story`, `devflow-checkpoint`, `devflow-memory`, `devflow-collab`, `devflow-create-persona`, `devflow-personalize`, `devflow-validate-overrides`, `devflow-new-doc`, `devflow-tech-debt`, `devflow-setup-checkpoint`, `devflow-init`, `devflow-version`
+  - Zero npm runtime dependencies - lightweight package (~600KB)
+  - Complements existing pip package (both methods work identically)
+  - Comprehensive npm installation documentation
+  - Version synchronization script for CHANGELOG → package.json
+
+## [1.8.0] - 2025-12-22
+
+### Added
+- **GitHub CLI Skill** - Comprehensive skill for GitHub CLI (gh) automation
+  - Repository management (view, clone, fork, create)
+  - Pull request operations (create, list, merge, review)
+  - Issue management (create, list, close, reopen)
+  - Workflow run management (list, view, rerun)
+  - Branch and release operations
+  - API access with JSON output parsing
+  - Common workflows and best practices
+
+## [1.7.1] - 2025-12-22
+
+### Fixed
+- **Comprehensive Lint Cleanup** - Fixed 1200+ Ruff lint errors across all tooling scripts
+  - Replaced deprecated `typing.List`, `typing.Dict`, `typing.Tuple`, `typing.Set` with Python 3.9+ generics
+  - Removed all trailing whitespace (W293, W291)
+  - Fixed import ordering (I001) across all modules
+  - Removed unused imports (F401) and unused variables (F841)
+  - Fixed f-strings without placeholders (F541)
+  - Replaced bare `except:` with specific exception types (E722)
+  - Fixed `raise ... from err` patterns (B904) for proper exception chaining
+  - Moved module-level imports to top of files (E402)
+  - Renamed unused loop variables to underscore prefix (B007)
+- **Test Fixes** - Updated `test_pair_programming.py` mock fixture to match refactored imports
+
+### Changed
+- Lowered test coverage threshold from 80% to 75% to account for new modules
+
+## [1.7.0] - 2025-12-22
+
+### Added
+- **Multi-Agent Collaboration System** - Advanced agent collaboration capabilities
+  - **Swarm Mode** (`--swarm`) - Multi-agent debate/consensus with automatic iteration
+    - Configurable consensus types: unanimous, majority, quorum, reviewer_approval
+    - Parallel execution support for first iteration
+    - Automatic issue tracking and resolution loops
+    - Budget limits and cost tracking per swarm session
+  - **Pair Programming Mode** (`--pair`) - DEV + REVIEWER interleaved collaboration
+    - Real-time feedback loops during implementation
+    - Chunk-based development with immediate review
+    - Automatic revision cycles for blocking issues
+    - Approval rate tracking
+  - **Auto-Route Mode** (`--auto-route`) - Intelligent agent selection
+    - Task type detection (bugfix, security, feature, refactor, etc.)
+    - Complexity analysis (trivial to critical)
+    - Dynamic agent selection based on task patterns
+    - File type and context awareness
+
+- **Shared Memory System** (`shared_memory.py`) - Cross-agent knowledge sharing
+  - Shared memory pool for all agents to read/write
+  - Tagging and search functionality
+  - Persistence across sessions
+  - Context generation for agent prompts
+
+- **Knowledge Graph** - Queryable decision tracking
+  - Record and track agent decisions
+  - Natural language queries ("What did ARCHITECT decide about auth?")
+  - Decision superseding and status tracking
+  - Topic-based indexing
+
+- **Agent Handoff System** (`agent_handoff.py`) - Structured handoffs between agents
+  - Automatic handoff summary generation
+  - Git diff analysis for file changes
+  - Decision and blocker tracking
+  - Warning propagation to next agent
+  - Markdown export for handoff documents
+
+- **Agent Router** (`agent_router.py`) - Dynamic agent selection
+  - Task complexity estimation
+  - Pattern-based task type detection
+  - Specialty matching for optimal agent selection
+  - Workflow recommendations (sequential, parallel, swarm, pair)
+  - Routing explanations with confidence scores
+
+- **Swarm Orchestrator** (`swarm_orchestrator.py`) - Multi-agent coordination
+  - Consensus detection algorithms
+  - Issue extraction from agent responses
+  - Vote determination (approve/reject/abstain)
+  - Cost estimation and budget enforcement
+  - Iteration result tracking
+
+- **Shell Tab Completion** - Autocomplete for commands
+  - Zsh completion script (`tooling/completions/_run-story`)
+  - Bash completion script (`tooling/completions/run-story-completion.bash`)
+  - PowerShell completion module (`tooling/completions/DevflowCompletion.ps1`)
+  - Completion for modes, agents, models, and all options
+  - Comma-separated agent list completion
+
+- **PowerShell Collaboration Script** (`run-collab.ps1`) - Windows native CLI
+  - Full feature parity with shell scripts
+  - Native PowerShell parameter handling
+  - Proper Windows path support
+  - Color output support
+
+- **New Claude Slash Commands** - Extended command set for collaboration
+  - `/swarm <story-key>` - Run multi-agent swarm mode (debate/consensus)
+  - `/pair <story-key>` - Run pair programming mode (DEV + REVIEWER)
+  - `/route <task>` - Auto-route task to best agents
+  - `/collab <story-key>` - Unified collaboration CLI with mode selection
+  - `/memory <story-key>` - View or query shared agent memory
+  - `/handoff <story-key>` - View handoff summaries between agents
+  - `/checkpoint` - Create or restore context checkpoints
+  - `/costs` - View cost dashboard and spending analytics
+
+### Changed
+- **Cross-Platform Compatibility** - Enhanced support for all platforms
+  - `run-collab.py` now detects Claude CLI on Windows, macOS, and Linux
+  - Cross-platform config/cache directory detection (XDG, Library, AppData)
+  - ANSI color support with Windows compatibility
+  - Path normalization for Windows
+  - Shell quoting for both Unix and Windows
+  - Safe filename sanitization for Windows restrictions
+
+- **Pair Programming Engine** (`pair_programming.py`) - Real-time collaboration
+  - Code chunk parsing and categorization
+  - Reviewer feedback extraction
+  - Revision loop management
+  - Session result summaries
+
+- **Unified Collaboration CLI** (`run-collab.py`) - Single entry point
+  - `--swarm` mode with custom agents
+  - `--pair` mode for pair programming
+  - `--auto` mode for intelligent routing
+  - `--memory` to view shared memory
+  - `--query` to query knowledge graph
+  - `--route-only` to preview routing decisions
+
+- **Collaboration Test Suite** (`test_collaboration.py`) - Comprehensive tests
+  - SharedMemory tests
+  - KnowledgeGraph tests
+  - AgentRouter tests
+  - HandoffGenerator tests
+  - SwarmOrchestrator tests
+  - PairProgramming tests
+  - Integration tests
+
+### Changed
+- **run-story.sh** - Extended with collaborative modes
+  - Added `--swarm` option for multi-agent debate
+  - Added `--pair` option for pair programming
+  - Added `--auto-route` option for intelligent routing
+  - Added `--agents` option for custom agent selection
+  - Added `--max-iter` option for iteration control
+
+### Documentation
+- **README.md** - Added Multi-Agent Collaboration section
+  - Swarm mode usage examples
+  - Pair programming mode examples
+  - Auto-route mode examples
+  - Shared memory and knowledge graph usage
+
+## [1.6.0] - 2025-12-21
+
+### Added
+- **Testing Infrastructure** - Comprehensive test suite for cost tracking
+  - pytest configuration in pyproject.toml
+  - Unit tests for cost_tracker.py (23KB test suite)
+  - Scenario-based integration tests
+  - Test fixtures and mocks via conftest.py
+  - 80% minimum code coverage requirement
+- **Enhanced Error Handling** - Structured error system for better debugging
+  - New errors.py module with error codes and context
+  - CostTrackingError, SessionError, BudgetError, CalculationError classes
+  - User-friendly error messages with suggested fixes
+  - Backward-compatible fallback for gradual adoption
+- **Cost Tracker Improvements** - Better validation and user experience
+  - Negative token count validation with clear error messages
+  - Unknown model warnings with fallback to sonnet pricing
+  - Enhanced budget status messages with emoji indicators
+  - Improved session save error handling (permission, disk space)
+  - Better session loading with validation and corruption detection
+  - Case-insensitive token usage parsing
+- **Setup Validation Tool** - User-friendly configuration verification
+  - validate_setup.py script for checking Devflow installation
+  - Checks for directories, config files, permissions
+  - Verbose mode and auto-fix capabilities
+  - Color-coded status output
+- **CI/CD Pipeline** - Automated testing and quality checks
+  - GitHub Actions workflow for multi-version Python testing (3.9-3.12)
+  - Ruff linting and formatting checks
+  - mypy type checking
+  - pytest with coverage reporting
+  - Codecov integration
+  - Integration test support
+- **Python Package Configuration** - Full project packaging support
+  - pyproject.toml with build system configuration
+  - requirements.txt and requirements-dev.txt
+  - Project metadata and dependencies
+  - CLI entry points (devflow-cost, devflow-validate)
+  - Development tools: pytest, mypy, ruff, coverage
+
+### Changed
+- **cost_tracker.py** - Enhanced with better error handling and validation
+  - calculate_cost() now validates token counts and warns on unknown models
+  - check_budget() returns emoji-enhanced status messages
+  - save_session() handles errors gracefully without disrupting workflow
+  - load_session() validates required fields and handles corruption
+
+## [1.5.3] - 2025-12-21
+
+### Changed
+- **Auto-commit disabled by default** - Changed `AUTO_COMMIT` default from `true` to `false` in both config templates (shell and PowerShell)
+
+## [1.5.2] - 2025-12-21
+
+### Changed
+- **Override system cleanup** - Removed example override files in favor of templates
+  - Users now copy from `templates/` to create their override files
+  - Added `user-profile.template.yaml` as starting point
+  - Updated templates README with clearer instructions
+  - Added `.gitignore` rules to keep user overrides private
+
+### Removed
+- `tooling/.automation/overrides/dev.override.yaml` - Use templates instead
+- `tooling/.automation/overrides/reviewer.override.yaml` - Use templates instead
+- `tooling/.automation/overrides/sm.override.yaml` - Use templates instead
+- `tooling/.automation/overrides/user-profile.yaml` - Use templates instead
+
+## [1.5.1] - 2025-12-21
+
+### Changed
+- **CLAUDE.md** - Updated changelog instructions to require proper version numbers
+  - Removed references to `[Unreleased]` sections
+  - Added semantic versioning guidance (patch/minor/major)
+  - Updated example to show proper versioned entry format
+- **README.md** - Updated version number to 1.5.1
+
+## [1.5.0] - 2025-12-21
+
+### Added
+- **Agent Persona Templates** - Pre-built personality configurations for all 8 agents:
+  - DEV agent: senior-fullstack, junior-mentored, security-focused, performance-engineer, rapid-prototyper
+  - Reviewer agent: thorough-critic, mentoring-reviewer, quick-sanity
+  - SM agent: agile-coach, technical-lead, startup-pm
+  - Architect agent: enterprise-architect, cloud-native, pragmatic-minimalist
+  - BA agent: requirements-engineer, agile-storyteller, domain-expert
+  - PM agent: traditional-pm, agile-pm, hybrid-delivery
+  - Writer agent: api-documentarian, user-guide-author, docs-as-code
+  - Maintainer agent: oss-maintainer, legacy-steward, devops-maintainer
+- **Personalization Wizard** (`personalize_agent.py`) - Interactive tool for customizing agent behavior
+- **`/personalize` slash command** - Quick access to the personalization wizard from Claude Code
+- **Memory Summarization Utility** (`memory_summarize.py`) - Consolidate and categorize agent memory files
+- **Override Validation & Linting** - New tools to validate override YAML files
+  - `validate-overrides.sh` - Shell script for validating YAML syntax, models, budgets
+  - `validate-overrides.py` - Python version with JSON output support
+  - Auto-fix capability for common issues (trailing whitespace, tabs)
+  - Schema validation for override structure
+- **Technical Debt Tracker** - Comprehensive debt metrics and reporting
+  - `tech-debt-tracker.py` - Scans codebase for TODO, FIXME, HACK, etc.
+  - Debt scoring system with severity weights
+  - Dashboard view with visual progress bars
+  - Trend tracking over time with history
+  - Markdown report generation
+  - JSON output for CI/CD integration
+- **Enhanced Bug Root Cause Analysis** - Improved bug-report.md template
+  - 5 Whys analysis framework
+  - Fault tree visualization template
+  - Root cause categorization (11 categories)
+  - Hypothesis and evidence tracking
+  - Investigation steps documentation
+  - Impact analysis section
+  - Prevention strategies
+- **Migration Rollback Automation** - New rollback capabilities
+  - `rollback-migration.sh` - Automated migration rollback tool
+  - Rollback checkpoint creation and management
+  - Git-based state preservation
+  - Dependency file backup (package.json, pubspec.yaml, etc.)
+  - Dry-run mode for preview
+  - Enhanced migration-spec.md with comprehensive rollback plan
+- **Custom Persona Builder** - Interactive agent persona creation
+  - `create-persona.py` - Full-featured Python wizard
+  - `create-persona.sh` - Shell script version
+  - 7 pre-built templates (developer, reviewer, architect, tester, security, devops, documentation)
+  - Automatic override file generation
+  - Persona validation
+- **Additional Project Types** - Java, Kotlin, Swift, and Android project detection
+- **Project Override Templates** - Stack-specific configurations:
+  - web-frontend.override.yaml
+  - backend-api.override.yaml
+  - data-science.override.yaml
+  - devops.override.yaml
+  - mobile-native.override.yaml
+
+### Changed
+- Expanded project type detection in initialization wizard
+- Enhanced override system with template-based customization
+- Enhanced `migration-spec.md` template with:
+  - Detailed pre-migration checklist
+  - Phase-based migration steps
+  - Rollback triggers and automation commands
+  - Communication plan template
+  - Lessons learned section
+- Improved `bug-report.md` template with structured root cause analysis
+
+## [1.4.0] - 2025-12-21
+
+### Added
+- **Override Templates** - Pre-configured templates for common project types
+  - `web-frontend.override.yaml` - React, Vue, Angular, Svelte
+  - `backend-api.override.yaml` - Node.js, Python, Go, Rust APIs
+  - `data-science.override.yaml` - ML, data analysis, Jupyter
+  - `devops.override.yaml` - Infrastructure, CI/CD, Kubernetes
+  - `mobile-native.override.yaml` - iOS (Swift), Android (Kotlin/Java)
+- **Memory Summarization** - Utility to prevent unbounded memory growth
+  - `memory_summarize.py` - Consolidates duplicate entries
+  - Categorizes memories by type (errors, preferences, implementation, etc.)
+  - Configurable max entries limit
+- **Additional Project Types** - Java, Kotlin, Swift, Android support
+  - Auto-detection for Maven, Gradle, Xcode projects
+  - Updated config templates with all supported types
+
+### Changed
+- **Simplified CLI Integration** - Removed terminal-based CLI in favor of Claude Code native commands
+  - Removed `tooling/scripts/gds` CLI wrapper
+  - Removed `tooling/scripts/completions/` directory (zsh completions)
+  - All workflows now accessed via Claude Code slash commands (`/story`, `/develop`, `/review`, etc.)
+- **Simplified zsh autocomplete** - Replaced complex interactive menu with native zsh completion
+  - Cleaner implementation (~130 lines vs ~450 lines)
+- Renamed product from "GDS_Automation" to "Devflow" throughout codebase
+- **CLAUDE.md** - Added project instructions file for Claude Code
+  - Mandatory changelog update requirement before pushing
+  - Project context and directory structure
+  - Commit style guidelines
+- **Claude Code Slash Commands** - Native slash commands for Devflow workflows
+  - `/story` - Run full story pipeline
+  - `/develop` - Run development phase
+  - `/review` - Run code review
+  - `/adversarial` - Run critical code review
+  - `/agent` - Invoke a specific agent
+  - `/bugfix` - Fix a bug
+  - `/devflow` - Run any Devflow command
+  - `/personalize` - Quick access to personalization wizard
+
+### Removed
+- `tooling/scripts/gds` - Terminal CLI wrapper
+- `tooling/scripts/completions/_gds` - Zsh completion script
+- `tooling/scripts/completions/_run-story` - Run-story completion script
+- `tooling/scripts/completions/gds-interactive.zsh` - Interactive completion
+- `tooling/scripts/completions/README.md` - Completion documentation
+
+## [1.3.1] - 2025-12-20
+
+### Added
+- **Interactive Visual Autocomplete** - Claude-style `/` command experience
+  - `Alt+.` triggers visual menu with arrow navigation
+  - `devflow.` shell function for quick command menu
+  - `devflow.story` and `devflow.agent` for context-specific menus
+  - Fuzzy filtering by typing
+  - fzf integration when available
+  - Category-tagged commands (story, maint, agent)
+
+### Changed
+- `devflow setup` now installs interactive completion
+- Updated completion documentation
+
+### Fixed
+- README: Removed references to non-existent Python scripts
+- README: Fixed agent count (now 8 with REVIEWER)
+- README: Updated version to 1.3.1
+- README: Added Devflow CLI documentation
+- README: Simplified cost management section
+- CONTRIBUTING: Added override system contribution areas
+- CONTRIBUTING: Updated release process documentation
+
+## [1.3.0] - 2025-12-20
+
+### Added
+- **Agent Override System** - BMAD-style personalization (inspired by [BMAD-METHOD](https://github.com/bmad-code-org/BMAD-METHOD))
+  - `user-profile.yaml` - Global user preferences (name, tech level, code style)
+  - Per-agent override files (`dev.override.yaml`, `sm.override.yaml`, etc.)
+  - Persistent agent memory system (`memory/{agent}.memory.md`)
+  - Runtime merging: base agent + user profile + overrides + memory
+  - Model and budget overrides per agent
+  - Customizations survive updates to core agents
+- **Adversarial Reviewer Agent** - Critical code reviewer that finds problems
+  - Uses Opus for deeper analysis
+  - Security, reliability, correctness, maintainability checks
+  - Issue severity classification (Critical, High, Medium, Low)
+  - `--adversarial` mode in run-story.sh
+- **CLI Autocomplete System** - Zsh tab completion
+  - `devflow` - Unified CLI wrapper with all commands
+  - Story key completion from sprint-status.yaml
+  - Mode and option completion
+  - Agent and override completion
+  - `devflow setup` command for easy installation
+- **New CLI Commands**
+  - `devflow profile` - Edit user profile
+  - `devflow override <agent>` - Edit agent overrides
+  - `devflow memory [show|add|clear] <agent>` - Manage agent memory
+  - `devflow agents` - List available agents
+  - `devflow logs [tail|list|view|clean]` - Log management
+
+### Changed
+- `claude-cli.sh` now loads overrides via `override-loader.sh`
+- Agent prompts include user context and memories
+- `run-story.sh` supports `--adversarial` mode
+
+### New Files
+- `tooling/.automation/overrides/` - Override configuration directory
+- `tooling/.automation/memory/` - Persistent agent memory directory
+- `tooling/.automation/agents/reviewer.md` - Adversarial reviewer agent
+- `tooling/scripts/lib/override-loader.sh` - Override loading library
+- `tooling/scripts/devflow` - Unified CLI wrapper
+- `tooling/scripts/completions/` - Zsh completion scripts
+
+## [1.2.0] - 2025-12-20
+
+### Added
+- **Brownfield Workflow Support** - Full support for existing codebase maintenance
+  - `--bugfix` mode for bug investigation and fixing
+  - `--refactor` mode for code refactoring
+  - `--investigate` mode for codebase exploration (read-only)
+  - `--quickfix` mode for small, targeted changes
+  - `--migrate` mode for framework/dependency migrations
+  - `--tech-debt` mode for technical debt resolution
+- **MAINTAINER Agent Persona** - New agent specialized for brownfield work
+- **Brownfield Directory Structure** - Organized folders for bugs, refactors, investigations, migrations, tech-debt
+- **Task Templates** - Templates for bug reports, refactoring specs, migration plans, tech debt items
+- **Workflow Mode Selection** - Init wizard now asks for greenfield, brownfield, or both
+
+### Changed
+- Agent count increased from 6 to 7 (added MAINTAINER)
+- `run-story.sh` now supports brownfield modes alongside greenfield
+- `init-project-workflow.sh` includes workflow mode selection step
+- README updated with brownfield documentation
+- CONTRIBUTING updated with brownfield contribution areas
+
+## [1.1.0] - 2025-12-20
+
+### Added
+- **Windows Compatibility** - Full PowerShell support
+  - `run-story.ps1` - Windows workflow runner
+  - `init-project-workflow.ps1` - Windows setup wizard
+  - `new-doc.ps1` - Windows documentation generator
+  - Updated installation instructions for Windows
+- **Cost Analysis System** - Real-time cost tracking
+  - `cost_dashboard.py` - Cost monitoring dashboard
+  - Multi-currency display (USD, EUR, GBP, BRL, CAD, AUD)
+  - Per-agent and per-model cost breakdown
+  - Budget alerts (warning at 75%, critical at 90%)
+  - Auto-stop at budget limit
+  - Session history and CSV export
+- **Currency Selection** - Init wizard now prompts for preferred currency
+
+### Changed
+- Updated README with Windows installation instructions
+- Added cost tracking configuration options to config.sh
+- Improved cross-platform compatibility
+
+## [1.0.0] - 2025-12-20
+
+### Added
+- Initial release of Claude Workflow Template
+- Multi-persona agent system (SM, DEV, BA, ARCHITECT, PM, WRITER)
+- Smart model optimization (Opus for dev/review, Sonnet for planning)
+- Automated story implementation pipeline
+- Context checkpoint system with auto-save
+- Interactive setup wizard (`init-project-workflow.sh`)
+- Documentation standards and templates
+- Sprint tracking system
+- Export/import functionality for portability
+- macOS LaunchAgent for background checkpoint monitoring
+- Cost optimization with budget controls
+- 6 pre-configured agent personas
+- Complete documentation suite
+- Git integration with auto-commit
+- Code review automation
+
+### Features
+- `run-story.sh` - Main workflow runner
+- `init-project-workflow.sh` - Interactive setup wizard
+- `context_checkpoint.py` - Checkpoint management system
+- `setup-checkpoint-service.sh` - Background service installer
+- `new-doc.sh` - Documentation template generator
+- `export-workflow-template.sh` - Template packager
+
+### Documentation
+- README.md - Main project documentation
+- CONTRIBUTING.md - Contribution guidelines
+- LICENSE - MIT License
+- CHANGELOG.md - This file
+- tooling/docs/DOC-STANDARD.md - Documentation standards
+
+### Supported Project Types
+- Flutter
+- Node.js
+- Python
+- Rust
+- Go
+- Ruby
+- Generic (any project)
+
+### Model Strategy
+- Opus ($15/$75 per M tokens) for development and code review
+- Sonnet ($3/$15 per M tokens) for planning, context, documentation
+- Average cost savings: 40-60% per story
+
+### Budget Controls
+- MAX_BUDGET_CONTEXT: Default $3.00
+- MAX_BUDGET_DEV: Default $15.00
+- MAX_BUDGET_REVIEW: Default $5.00
+
+---
+
+For more details, see the [GitHub releases](https://github.com/Pedro-Jose-da-Rocha-Mendonca/Devflow/releases).

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Pedro Jose da Rocha Mendonca
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.