Skip to content

Latest commit

 

History

History
259 lines (214 loc) · 12.3 KB

File metadata and controls

259 lines (214 loc) · 12.3 KB

Changelog

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.


[1.5.0] - 2025-11-06

Added

  • 3 Production-Ready Skills:
    • Scrum Master Agent (15 files) - Sprint analytics, backlog prioritization, multi-tool integration
    • TDD Guide (15 files) - Test generation, coverage analysis, multi-framework support
    • Tech Stack Evaluator (14 files) - TCO calculation, security assessment, migration analysis
  • GitHub CI/CD Specialist Mega-Prompt (~4,800 tokens) - Complete GitHub Actions workflow blueprint
  • Git Governance Workflows:
    • Commit validation (conventional commits with commitlint)
    • Branch naming enforcement (feat/, fix/, docs/*, etc.)
    • CI quality gates (Python syntax, markdown links, security audit)
    • Release orchestration
  • 10 Git Slash Commands for workflow automation

Changed

  • Disabled yamllint check in CI (creates false positives with markdown in bash strings)
  • Updated SKILLS_FACTORY_PROMPT with file cleanliness standards

Fixed

  • Branch naming convention validation (slash instead of dash)
  • YAML syntax in workflow files (heredoc format for multiline strings)

[1.4.0] - 2025-10-30

Added

  • Codex CLI Bridge skill (48 KB) - Seamless interoperability between Claude Code and OpenAI Codex CLI
    • Reference-based architecture (no file duplication)
    • Automatic CLAUDE.md → AGENTS.md translation
    • Cross-tool team collaboration support
  • 7 Python modules for bridge functionality:
    • bridge.py - Main orchestrator with CLI interface
    • safety_mechanism.py - Environment validation (Codex CLI + CLAUDE.md checks)
    • claude_parser.py - CLAUDE.md parser with YAML frontmatter support
    • project_analyzer.py - Project structure and metadata analysis
    • agents_md_generator.py - Template-based AGENTS.md generation
    • skill_documenter.py - Different approaches for functional vs prompt-based skills
    • codex_executor.py - Codex CLI helpers (always uses 'codex exec')
  • 2 new slash commands:
    • /sync-agents-md - Regenerate AGENTS.md from CLAUDE.md (integrates with /init, /update-claude, /check-docs)
    • /codex-exec - Execute Codex CLI commands with proper syntax (analysis, edit, search, resume types)
  • AGENTS.md generation - Auto-generates Codex CLI-compatible documentation from CLAUDE.md
    • 19 KB output for claude-code-skills-factory (629 lines, 13 skills, 59 agents)
    • Workflow pattern translations (Claude commands → Codex equivalents)
    • Command reference table (Claude Code ↔ Codex CLI)
    • MCP integration documentation
  • Safety mechanisms:
    • Auto-validates Codex CLI installation
    • Auto-runs /init if CLAUDE.md missing (with user notification)
    • Minimal CLAUDE.md fallback generation
    • Comprehensive error handling
  • Package distribution - package.sh script creates distributable ZIP (48 KB) with SHA256 checksum

Changed

  • README.md completely revised for better focus on shortcuts and efficiency
    • Added "Quick Start (3 Shortcuts)" section at top
    • Added "Built-in Commands" section with tables for 7 slash commands and 4 interactive agents
    • Added "Complete Workflow Examples" with 3 real-world scenarios
    • Enhanced Repository Structure with detailed .claude/ breakdown
    • Improved Production Skills section with "Most Popular" and "New" badges
    • Added emojis for better scannability
    • Reduced from 876 to 367 lines while adding more practical content
  • Repository structure - Added AGENTS.md to root and new slash commands to .claude/commands/

Documentation

  • Use Cases: Cross-tool teams, project migration, CI/CD integration, skills marketplace compatibility
  • Model selection helpers: gpt-5 (general reasoning) vs gpt-5-codex (code editing)
  • Sandbox modes: read-only, workspace-write, danger-full-access
  • Workflow integration: /sync-agents-md integrates with existing Claude Code commands
  • Complete documentation (6 files): SKILL.md, README, HOW_TO_USE, INSTALL, CHANGELOG, samples

Fixed

  • Consolidated README.md for 58% reduction in length (876 → 367 lines) while adding practical shortcuts
  • Better focus on built-in commands and workflow efficiency

[2.2.0] - 2025-10-29

Added

  • MASTER_SLASH_COMMANDS_PROMPT.md - Comprehensive master template (~1500 lines) for generating slash commands following official Anthropic patterns
  • documentation/references/ folder - 6 official Anthropic slash command examples for pattern reference
  • Three official command structure patterns integrated into slash-command-factory:
    • Simple Pattern (code-review) - Context → Task for straightforward workflows
    • Multi-Phase Pattern (codebase-analyze) - Discovery → Analysis → Task for complex documentation
    • Agent-Style Pattern (ultrathink, openapi-sync) - Role → Process → Guidelines for expert coordination
  • 7 new official presets - code-review, codebase-analyze, update-docs, openapi-sync, ultrathink, deps-audit, metrics-report
  • Comprehensive four-layer validation - Command name (kebab-case), bash permissions (specific only), arguments usage ($ARGUMENTS), YAML structure
  • Smart generation features:
    • Auto-detect command structure pattern based on purpose
    • Auto-generate bash permissions (git, discovery, comprehensive patterns)
    • Automatic kebab-case naming conversion with validation
  • Test command - marketing-research.md generated and validated as demonstration

Changed

  • Slash Command Factory enhanced to v2.0 (26 KB)
    • Expanded from 10 to 17 presets (10 original + 7 official examples)
    • SKILL.md enhanced with official patterns documentation (~1000 lines)
    • command_generator.py enhanced with structure detection, naming validation, bash permission generation
    • validator.py enhanced with comprehensive four-layer validation system
    • presets.json expanded with 7 new official preset commands
    • HOW_TO_USE.md updated with pattern guide and naming convention examples
  • README.md updated to version 1.3.0
    • Repository structure diagram includes all 4 templates and documentation/references/
    • Slash Command Factory section enhanced with v2.0 details
    • Quick Reference updated with MASTER_SLASH_COMMANDS_PROMPT.md
    • Version section updated with comprehensive v1.3.0 changelog
  • CLAUDE.md updated
    • Repository structure diagram includes all 4 templates and documentation/references/
    • Templates for Generation section enhanced with MASTER_SLASH_COMMANDS_PROMPT.md details
    • Slash Command Factory (#11) description enhanced with comprehensive v2.0 capabilities

Documentation

  • Validation strictness - All generated commands must pass four-layer validation
  • Bash permission patterns - Never uses wildcard Bash, always specific commands (Bash(git status:*))
  • Naming convention - Enforced kebab-case (2-4 words) with automatic conversion algorithm
  • Arguments standard - Always $ARGUMENTS, never positional $1, $2, $3

[2.1.0] - 2025-10-29

Added

  • Slash Command Factory skill - Generate custom Claude Code slash commands with 10 powerful presets
  • Interactive Navigation Agents (4 agents) - Conversational guides for building Skills, Prompts, Agents, and Commands
    • factory-guide (orchestrator) - Delegates to appropriate specialist
    • skills-guide - Builds Claude Skills through 4-5 questions
    • prompts-guide - Guides prompt-factory usage (69 presets)
    • agents-guide - Builds Claude Code Agents through 5-6 questions
  • 5 Workflow Slash Commands - Complete start-to-finish workflow
    • /build - Start building (invokes orchestrator or specialists)
    • /validate-output - Quality check with auto-ZIP creation
    • /install-skill - Installation guidance
    • /test-factory - Testing examples
    • /factory-status - Progress tracking
  • Public Gist - Slash Command Factory guide for SEO/AEO
  • 10 Slash Command Presets - Business research, content analysis, medical translation, compliance audit, API building, test automation, docs generation, knowledge extraction, workflow optimization, agent coordination

Changed

  • README reorganized - Navigation system moved to end (before Contributing)
  • README enhanced - Added Slash Command Factory (#8 in Generated Skills)
  • /validate-output - Now auto-creates ZIP files on successful validation
  • CLAUDE.md - Added Slash Command Factory (#11) documentation
  • GISTS.md - Added 4th public gist (Slash Command Factory)
  • Documentation references - Updated with slash-command-factory links

Added Documentation

  • .claude/agents/ - 4 navigation agents (orchestrator + 3 specialists)
  • .claude/commands/ - 5 workflow commands
  • generated-skills/slash-command-factory/ - Complete skill with presets
  • documentation/references/ - Slash command examples from Anthropic

[2.0.0] - 2025-10-28

Added

  • Health SDK Builder skill - Comprehensive healthcare Agent SDK builder with HIPAA/GDPR/DSGVO/PTV 10 compliance
  • Multilingual support - German + English mandatory for healthcare applications
  • Therapy modalities - CBT, Psychodynamic, Psychoanalysis, Depth Psychology
  • German PTV 10 automation - 60-session therapy applications (2.5 DIN A4 pages)
  • Medical terminology translation - 8th-10th grade reading level for patients
  • 2025 Claude API features - Computer Use, Code Execution, Files API, Extended Caching
  • PR-to-issue auto-close workflow - Automatically close linked issues when PRs merge
  • README Option C - Prompt Factory as third quick-start path
  • Public Gists - 3 comprehensive guides for SEO and Answer Engine Optimization
  • Professional standards - CONTRIBUTING.md, LICENSE (MIT), .editorconfig
  • .github consolidation - 6 files merged into single GITHUB_WORKFLOWS_GUIDE.md
  • .archive structure - Organized historical documents
  • Security audit - Complete security audit and SECURITY.md policy

Changed

  • README enhanced - Added Complementary Resources (Claude Code Tresor, Claude Skills Library)
  • README structure - Three-path quick start (Skills, Agents, Prompts)
  • prompt-suite → prompt-factory - Renamed for clarity
  • Repository size - Reduced from 37MB to 10MB (73% reduction)
  • .gitignore updated - Proper exclusions for test/development content

Removed

  • test-workspace/ - Moved to separate testing repository (31MB freed)
  • github-docs/ - Temporary reference files deleted (48KB)
  • .DS_Store files - macOS metadata cleaned (7 files)
  • claude-skills-examples/ from tracking - Reference material, gitignored
  • Outdated workflow docs - 6 files consolidated into 1
  • Internal files from tracking - TEST_CLAUDE_REVIEW.md, claude-skills-instructions.md

Fixed

  • Skill naming - health-sdk-builder (removed "claude" reserved word)
  • Gitignore rules - Proper exclusions for all test/dev content
  • Documentation references - Updated to point to consolidated guides

Security

  • API key isolation - Separate testing repository for SDK development
  • Gitignore enhanced - All sensitive files excluded
  • .archive/ - Internal documents organized and excluded

[1.2.0] - 2025-10-23

Added

  • Prompt Factory skill (403KB) - World-class prompt generation with 69 presets
  • PROMPTS_FACTORY_PROMPT.md template
  • Generated prompts examples (5 sample outputs)
  • Marketing growth prompt builder

Changed

  • Expanded coverage to 15 professional domains
  • Multi-format output (XML/Claude/ChatGPT/Gemini)

[1.1.0] - 2025-10-21

Added

  • Psychology Advisor skill (31KB) - Evidence-based mental wellness
  • CBT techniques (cognitive distortions, thought records)
  • Mindfulness exercises
  • Stress management tools

[1.0.0] - 2025-10-20

Added

  • Initial repository structure
  • SKILLS_FACTORY_PROMPT.md template
  • AGENTS_FACTORY_PROMPT.md template
  • Example skills (Financial Analysis, Brand Guidelines, Financial Models)
  • AWS Solution Architect skill (53KB)
  • Content Trend Researcher skill (35KB)
  • Microsoft 365 Tenant Manager skill (40KB)
  • Agent Factory skill (12KB)
  • GitHub automation workflows (10 workflows)
  • Documentation structure

Version Guidelines

Major version (X.0.0): Breaking changes, major new features Minor version (0.X.0): New features, backward compatible Patch version (0.0.X): Bug fixes, documentation updates


Current Version: 1.4.0 Last Updated: October 30, 2025