- Migrate all 48 Opus agents to Sonnet
- Optimize 35 execution-focused agents for Haiku
- Update README with hybrid orchestration patterns
- Simplify model configuration to use agnostic aliases
Final distribution: 97 Sonnet / 47 Haiku agents
- Revert plugin name from embedded-systems to arm-cortex-microcontrollers
- Keep original name chosen by contributor Ryan Snodgrass
- Update all references in README.md and marketplace.json
- Plugin structure remains properly organized in plugins/ directory
- Refactor standalone agent into proper plugin structure
- Create plugins/embedded-systems/ directory with agents/ subdirectory
- Move arm-cortex-expert.md to plugins/embedded-systems/agents/
- Update marketplace.json plugin name and source path
- Update README.md to reflect 7 language plugins (was 6)
- Add embedded-systems to Languages plugin table
- Update all agent references to new path
- Remove empty agents/ directory
This aligns the embedded-systems plugin with the standard plugin
architecture used throughout the repository.
Updated all 84 agent references from incorrect `agents/{agent-name}.md` format to correct `plugins/{plugin-name}/agents/{agent-name}.md` paths to resolve 404 errors.
- Add comprehensive slash command documentation as primary interface
- Include real command examples mapped to actual plugins in marketplace
- Add 70+ categorized slash command reference (all real commands)
- Show both command-based and natural language invocation patterns
- Add tdd-workflows to Testing section for better discoverability
- Provide detailed workflow orchestration examples with step breakdowns
- Include command argument examples and discovery instructions
All examples now use actual commands that exist in the plugin system
(backend-development:feature-development, unit-testing:test-generate, etc.)
- Organize 62 plugins into isolated directories under plugins/
- Consolidate tools and workflows into commands/ following Anthropic conventions
- Update marketplace.json with isolated source paths for each plugin
- Revise README to reflect plugin-based structure and token efficiency
- Remove shared resource directories (agents/, tools/, workflows/)
Each plugin now contains only its specific agents and commands, enabling
granular installation and minimal token usage. Installing a single plugin
loads only its resources rather than the entire marketplace.
Structure: plugins/{plugin-name}/{agents/,commands/}
- Update terraform-specialist from opus to sonnet
- Update database-optimizer from opus to sonnet
- Update tdd-orchestrator from opus to sonnet
- Update tutorial-engineer from opus to sonnet
- Update README.md with new model distribution (Sonnet: 52, Opus: 20)
- Remove duplicate agents/README.md file
These agents are implementation-focused rather than reasoning-focused,
making Sonnet the more appropriate model choice. This change follows
feedback from PR #72 while maintaining our task-complexity-tier
philosophy.
- Update primary installation method to use full HTTPS URL
- Remove redundant manual installation section
- Simplify instructions to two commands
- Resolves SSH authentication issues reported in #85
Users were experiencing "Failed to clone marketplace repository" errors
when using the short format (wshobson/agents) because Claude Code
defaults to SSH cloning. Using the full HTTPS URL bypasses this issue.
- Replace Quick Start section with interactive /plugin command instructions
- Add note about using wshobson/agents as marketplace source
- Simplify installation steps to match actual UI prompts
- Keep Available Plugins section as reference
Major Updates:
- Add Plugin Marketplace as primary installation method (Available Oct 9, 2025)
- Show /plugin commands for installing workflow-based collections
- Reorganize structure: Installation → Agent Categories → Model Config → Usage
- Remove Migration section (not relevant for this PR)
- Fix heading hierarchy issues from merge
- Preserve all 83 agent details in comprehensive tables
- Add workflow and tool invocation examples with proper syntax
The README now highlights the new plugin marketplace feature while
maintaining complete agent documentation and technical specifications.
The migration guide referenced slash command syntax that won't work
until the plugin system is deployed. Removed to avoid user confusion.
The repository now focuses on what works today: 83 agents.
Reverted README to highlight the 83 agents which are the core feature.
Added small note about repository update with link to MIGRATION.md.
Updated all agent links to point to agents/ subdirectory.
The agents work now and are the primary value. Workflows/tools will
be documented in the plugin PR tomorrow.
Workflows and tools will be accessible via plugin system (coming soon).
For now, the README focuses on the 83 agents which are directly usable.
Changes:
- Removed /workflows: and /tools: invocation examples
- Updated multi-agent orchestration examples to show natural language
- Simplified migration guide to focus on what works now
- Plugin-based workflow/tool access will be documented in future PR
Documentation Updates:
- Add migration warning and quick links section to README
- Expand installation section with update instructions
- Create comprehensive MIGRATION.md guide with:
* Step-by-step migration process
* Before/after structure comparison
* Command syntax changes reference
* Common issues and solutions
* Testing instructions and rollback steps
- Document breaking changes in command invocation syntax
- Add troubleshooting resources and help links
This provides a smooth user experience for the major repository restructure.
Repository Restructure:
- Move all 83 agent .md files to agents/ subdirectory
- Add 15 workflow orchestrators from commands repo to workflows/
- Add 42 development tools from commands repo to tools/
- Update README for unified repository structure
The commands repository functionality is now fully integrated, providing
complete workflow orchestration and development tooling alongside agents.
Directory Structure:
- agents/ - 83 specialized AI agents
- workflows/ - 15 multi-agent orchestration commands
- tools/ - 42 focused development utilities
No breaking changes to agent functionality - all agents remain accessible
with same names and behavior. Adds workflow and tool commands for enhanced
multi-agent coordination capabilities.
Repository Restructure:
- Move all 83 agent .md files to agents/ subdirectory
- Add 15 workflow orchestrators from commands repo to workflows/
- Add 42 development tools from commands repo to tools/
- Update README for unified repository structure
This prepares the repository for unified plugin marketplace integration.
The commands repository functionality is now fully integrated, providing
complete workflow orchestration and development tooling alongside agents.
Directory Structure:
- agents/ - 83 specialized AI agents
- workflows/ - 15 multi-agent orchestration commands
- tools/ - 42 focused development utilities
No breaking changes to agent functionality - all agents remain accessible
with same names and behavior. Adds workflow and tool commands for enhanced
multi-agent coordination capabilities.
- Added observability-engineer to Performance & Observability section
- Updated total agent count from 82 to 83
- Renamed "Performance & Research" to "Performance & Observability"
- Updated Opus model count from 21 to 22 agents
- Added observability-engineer to Critical Analysis category in Opus section
- Updated agent count from 78 to 82
- Added 3 new security coding agents (backend, frontend, mobile) to Code Quality & Review section
- Added blockchain-developer agent to Specialized Domains section
- All new agents properly linked with descriptions
- Convert all agent listings to professional tables
- Add structured tables with Agent, Model, and Description columns
- Reorganize model configuration section with distribution tables
- Enhance agent selection guide with task-based tables
- Maintain consistency with commands repository formatting
- Improve readability and information accessibility
- Keep all existing content while improving presentation
- Remove all emoji for cleaner professional appearance
- Reorganize content with logical categorical structure
- Improve technical clarity and concise descriptions
- Add clear hierarchy with consistent formatting
- Focus on functionality rather than marketing language
- Better organization of agents by functional domains
- Streamline model configuration section
- Enhance usage examples with technical focus
- Reduce content from 558 to 438 lines while maintaining all essential information
- Add new tdd-orchestrator agent for managing complete TDD workflows
- Enhance test-automator with dedicated TDD capabilities and metrics
- Add TDD excellence section with red-green-refactor cycle automation
- Include support for Chicago and London TDD schools
- Add property-based TDD approaches and BDD integration
- Include TDD cycle metrics tracking and compliance monitoring
- Update README with new agent and TDD examples
- Add comprehensive TDD usage examples and documentation
This enhancement provides full TDD orchestration capabilities across
the agent collection, enabling test-first development with automated
cycle management, metrics tracking, and multi-agent coordination.
Addresses issue #64: TDD workflow support
- Enhanced all agents with 2024/2025 best practices and modern tooling
- Standardized format with 8-12 detailed capability subsections per agent
- Added Django Pro and FastAPI Pro specialist agents
- Updated model assignments (Sonnet/Haiku) based on task complexity
- Integrated latest frameworks: React 19, Next.js 15, Flutter 3.x, Unity 6, etc.
- Enhanced infrastructure agents with GitOps, OpenTelemetry, service mesh
- Modernized AI/ML agents with LLM integration, RAG systems, vector databases
- Updated business agents with AI-powered tools and automation
- Refreshed all programming language agents with current ecosystem tools
- Enhanced documentation with comprehensive README reflecting all improvements
Total changes: 5,945 insertions, 1,443 deletions across 40 files
All agents now provide production-ready, enterprise-level expertise
- Update agent count from 71 to 75 to reflect current repository state
- Correct model assignment counts: Haiku (15), Sonnet (45), Opus (15)
- Add hr-pro to Sonnet model section and Business & Marketing guidance
- Fix model distribution after recent PR merges
- Add comprehensive Flutter development agent with expertise in Dart, widgets, and cross-platform features
- Supports state management patterns (Provider, Riverpod, Bloc, GetX)
- Handles platform-specific integrations for iOS, Android, Web, and desktop
- Includes performance optimization, testing strategies, and deployment configuration
- Update README to reflect 61 total agents (37 Sonnet agents)
