Bundled plugins for Claude Code to orchestrate coding agents.
npx claudepluginhub seekayel/agent-skillsRalph plugin for Claude Code
Structured workflow for AI-assisted development using research, planning, and systematic implementation phases
Code review skill using the codex CLI
Claude Code marketplace entries for the plugin-safe Antigravity Awesome Skills library and its compatible editorial bundles.
Production-ready workflow orchestration with 79 focused plugins, 184 specialized agents, and 150 skills - optimized for granular installation and minimal token usage
Directory of popular Claude Code extensions including development tools, productivity plugins, and MCP integrations
Collection of skills for Claude Code agents.
A structured workflow for AI-assisted development using research, planning, and systematic implementation phases. Includes specialized sub-agents for codebase exploration.
Tools for orchestrating and improving coding agent workflows.
To use these skills in your project, add the following to your local AGENTS.md file (or CLAUDE.md):
Add this snippet to ensure Claude uses the structured research-plan-implement approach for all code changes:
## Development Workflow
When implementing changes, features, or bug fixes, follow the research-plan-implement workflow:
### Before Making Changes
1. **Research Phase**: Before modifying any code, thoroughly research the codebase:
- Use the `codebase-locator` agent to find relevant files and directories
- Use the `codebase-analyzer` agent to understand code structure and dependencies
- Use the `pattern-finder` agent to identify existing patterns and conventions
- Document findings before proceeding
2. **Planning Phase**: Create a detailed implementation plan:
- List all files that need to be created or modified
- Break work into logical phases with clear success criteria
- Identify dependencies between changes
- Define testing strategy
3. **Validation Phase**: Before implementing, verify:
- All necessary files have been identified
- Changes follow existing patterns
- Success criteria are testable
4. **Implementation Phase**: Execute systematically:
- Follow the plan in phase order
- Verify success criteria at each step
- Commit after completing each phase
5. **Testing Phase**: Run the tests defined in the Planning phase:
- Execute all test cases from the testing strategy
- Verify all success criteria are met
- Fix any failures before considering the work complete
- Do not skip tests or mark work as done until all tests pass
### Agent Usage
When researching the codebase, spawn these specialized agents in parallel:
- **codebase-locator**: "Find all files related to [feature/component]"
- **codebase-analyzer**: "Analyze the architecture of [system/module]"
- **pattern-finder**: "Identify patterns used for [type of code]"
### Required Outputs
For any non-trivial change, produce:
1. Research summary documenting relevant files and patterns found
2. Implementation plan with phases and success criteria
3. Validation checklist confirming plan completeness
For a more concise configuration, use this minimal snippet:
## Development Workflow
For all code changes:
1. **Research first** - Use codebase-locator, codebase-analyzer, and pattern-finder agents to understand the codebase before making changes
2. **Plan before implementing** - Create a phased implementation plan with success criteria and testing strategy
3. **Validate the plan** - Verify completeness before writing code
4. **Implement systematically** - Follow the plan, verify each phase, commit often
5. **Test thoroughly** - Run all tests from the plan, fix failures, do not skip tests
If you want to emphasize parallel agent usage:
## Agent-Driven Research
Before implementing any feature or fix, spawn these research agents in parallel:
| Agent | Purpose | Example Prompt |
|-------|---------|----------------|
| codebase-locator | Find relevant files | "Locate all authentication-related files" |
| codebase-analyzer | Understand structure | "Analyze how the API layer handles requests" |
| pattern-finder | Identify conventions | "Find patterns for creating new endpoints" |
Wait for all agents to complete, then synthesize findings into an implementation plan before writing any code.
Add the agent-skills marketplace to your Claude Code configuration:
# Add the marketplace to your global Claude Code settings
claude mcp add agent-skills-marketplace https://github.com/seekayel/agent-skills
Or manually add to your ~/.claude/settings.json:
{
"plugins": {
"marketplaces": [
{
"name": "agent-skills-marketplace",
"source": "https://github.com/seekayel/agent-skills"
}
]
}
}
Once the marketplace is added, install specific plugins:
# Install the research-plan-implement plugin
claude plugin install research-plan-implement-plugin
# Install the orchestration plugin
claude plugin install orchestration-plugin
Or add to your project's .claude/settings.json:
{
"plugins": {
"installed": [
"agent-skills-marketplace/research-plan-implement-plugin",
"agent-skills-marketplace/orchestration-plugin"
]
}
}